REPORT NO , REV. 1

Transcription

1 METALLURGICAL INVESTIGATION OF A FRACTURED SECTION OF THE 20" O.D. PIPELINE AT MILEPOST IN THE CONWAY TO CORSICANA SEGMENT OF THE PEGASUS CRUDE OIL PIPELINE REPORT NO , REV. 1 Prepared for ExxonMobil Pipeline Company and the Pipeline and Hazardous Materials Safety Administration pursuant to Corrective Action Order CPF H

2 TABLE OF CONTENTS Page i 1.0 INTRODUCTION 1.1 Brief Narrative of the Incident 1.2 Scope of the Investigation 1.3 Development of Test Protocol 2.0 BACKGROUND INFORMATION 2.1 Pipe Manufacturing and Coating 2.2 Inspection and Service History 2.3 Specifications 2.4 Items Received for Testing 3.0 METALLURGICAL EXAMINATION, TESTING, AND ANALYSIS 3.1 Visual and Macroscopic Observations 3.2 As-Received Condition of the Pipe and Coating 3.3 Coating Removal Process 3.4 Condition of the Pipe Following Coating Removal 3.5 Dimensional Measurements 3.6 Residual Stresses 3.7 Fractographic Examinations 3.8 Crack Measurements 3.9 Metallographic Evaluation 3.10 Microhardness Surveys 3.11 Tensile Tests 3.12 Charpy V-Notch Impact Tests 3.13 Chemical Analyses 4.0 CONCLUSION 4.1 Technical Causes of Failure 4.2 Failure Scenario

3 Page ii TABLES Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15 Table 16 Table 17 Table 18 Table 19 Table 20 Out-of-Roundness Measurements/Calculations Wall Thickness Measurements along Fracture Surface Hook Crack(s) Depth Crack Width Estimates Microhardness Survey at Fractured Area Microhardness Survey at Fractured Area Microhardness Survey at Intact Area Tensile Test - ERW Transverse Tensile Test - Base Metal Transverse Tensile Test - Base Metal Longitudinal Tensile Test - Sub-sized Round Transverse Charpy V-notch Impact Test - ERW Transverse Charpy V-notch Impact Test - Heat-Affected Zone (HAZ) Transverse Charpy V-notch Impact Test - Base Metal Transverse Chemical Analysis - OES Base Metal Chemical Analysis - EDS Fracture Surface Chemical Analysis - EDS Fracture Surface Chemical Analysis - EDS Fracture Surface Chemical Analysis - EDS O.D. Corrosion Chemical Analysis - EDS O.D. Bitumen Coating APPENDICES Appendix I Test Protocol Appendix II Chain of Custody Appendix III Coating Removal Photographs and Documents Appendix IV UT Wall Thickness Results Appendix V Location of Specimen Removal

4 METALLURGICAL INVESTIGATION OF A FRACTURED SECTION OF THE 20" O.D. PIPELINE AT MILEPOST IN THE CONWAY TO CORSICANA SEGMENT OF THE PEGASUS CRUDE OIL PIPELINE 1.0 INTRODUCTION 1.1 Brief Narrative of the Incident On March 29, 2013 at 2:37 pm CST, a drop in pressure was detected within the Pegasus Pipeline of the Conway to Corsicana line segment by ExxonMobil Pipeline Company (EMPCo) at their Operations Control Center in Houston, Texas. The cause of the pressure drop was the rupture of a section of the pipeline at Milepost in Mayflower, Arkansas. The operating pressure at the time of failure was estimated to be between 702 psig and 708 psig. 1.2 Scope of the Investigation Hurst Metallurgical Research Laboratory, Inc. (HurstLab) was retained by EMPCo, with approval by the U.S. Department of Transportation, Pipeline and Hazardous Materials Safety Administration (PHMSA), to provide technical support in the investigation of the failed section of the pipeline, as well as conduct and direct the required metallurgical tests to determine, if possible, the root cause of the failure, pursuant to Corrective Action Order CPF H. The investigation of the cracked section of the pipeline conducted by HurstLab is a joint effort by various staff members of the Laboratory, which includes some of the report writing and analysis conducted by Susan Dalrymple-Ely, Materials Analyst and metallurgical tests conducted by Clint Myers, Staff Metallurgist of the Laboratory. The investigative effort made by this Laboratory also includes a review of the UT data and SEM fractographs provided by approved vendors.

5 Page 2 of 185 The investigation conducted by this Laboratory is primarily based on the tests and analyses performed in accordance with the approved test protocol, review of the available information, and research conducted by this Laboratory. We reserve the right to change, amend, or omit our opinions, as warranted, based upon any additional information or further test results that may be obtained or made available to this Laboratory. 1.3 Development of Test Protocol On April 13, 2013, a preliminary metallurgical test protocol was development by HurstLab following the general guideline entitled Metallurgical Laboratory Examination Protocol dated 05/08/2007 for metallurgical failure investigation of pipeline prepared by PHMSA. Following various revisions that were made to incorporate the changes requested by PHMSA, a protocol entitled Pegasus Line - Conway to Corsicana M.P , Mechanical and Metallurgical Testing and Failure Analysis Protocol, referenced as Test Protocol Rev. 4, CPF No H, Amended 4/18/13, was developed and was approved by PHMSA. A copy of the final approved protocol is presented in Appendix I. 2.0 BACKGROUND INFORMATION 2.1 Pipe Manufacturing and Coating The subject section of the 20" Patoka to Corsicana #1-20" North Pipeline, the segment from Conway to Corsicana, consisted of approximately 50' long sections of 20" O.D. x 0.312" thick wall DC Electric Resistance Welded (ERW) pipe that was manufactured in 1947 and 1948 by Youngstown Sheet and Tube Company in Youngstown, Ohio. The welded pipe was manufactured from Open Hearth Steel meeting Grade B mechanical requirements The O.D. surface of the pipeline was coated with some type of a viscous bitumen or coal-tar coating, on top of which was a layer of somewhat harder but more brittle fibrous coating. No details concerning the coating type or process were available. The pipeline had reportedly been impressed current cathodically protected since installation, with possible anodes as well. The weight of the coated pipe was reported to be lbf/ft.

6 Page 3 of Inspection and Service History The subject section of pipeline was placed in service in 1948, and was buried approximately 3' below ground in native sandy clay soil. The pipeline carried crude oil from west Texas to Patoka, Illinois between 1948 and From 1995 to 2002 the line carried both west Texas crude oil and foreign crude oil (via the Gulf of Mexico) northward. In December 2002 the line was purged and idled with nitrogen. The pipeline containing the subject section of the pipe was successfully hydrostatic tested on January 24, 2006 at 1082 psig, which established a calculated MAOP of 866 psig at the failure location, based upon the Arkansas River ROV test site pressure at 1091 psig adjusted for elevation difference to the failure location. The line was then placed back in service transporting crude oil south towards the Gulf of Mexico, and remained in service up until the time of the failure Prior to failure, the pipeline was reported to typically operate between 47 F and 78 at pressures ranging between 240 psig and 820 psig. The pressure at the time of the failure was estimated to be between 702 psig and 708 psig. The fractured segment of the pipeline was located in a cleared right-of-way at the edge of a subdivision. No trees, roads, or buildings were located directly above the pipeline where the fracture occurred. As shown in Photograph No. 1, two (2) homes were built in close proximity to the pipeline, with driveways crossing over the pipeline at two (2) points downstream of the fractured segment. During construction of the homes, the pipeline may have experienced vehicle loadings caused by construction equipment and/or vehicles crossing the pipeline at multiple locations, including over the fractured segment. There was no indication of construction, digging, localized flooding, or other ground movements in the area of the fractured segment occurring during or immediately prior to the pipeline rupture. 2.3 Specifications At the request of EMPCo, the subject pipe was compared to two (2) versions of the API 5L specification throughout this report, both the edition that was in effect at the time the pipe was manufactured, and the current edition of said specification, both of which are detailed below.

7 Page 4 of At the time the pipe was manufactured in 1947 and 1948, the th specification in effect was API STD. 5-L, 10 Edition, August Per this specification, the smelting type of steel was reportedly Open Hearth Steel, the pipe was classified as an Electric Welded Pipe, and the strength was specified to meet Grade B requirements. This edition will be th referred to as API 5-L, 10 Edition throughout the report and the accompanying tables The currently applicable edition of the specification is ANSI/API 5L, 44 th Edition, Effective October 1, 2007, with Errata dated January 2009, Addendum 1 dated February 2009, Addendum 2 dated April 2010, and Addendum 3 dated July The requirements for PSL 1 Welded Pipe, Grade X42 will be used for comparison, with the exception of the Charpy V-Notch (CVN) impact tests. For the CVN impact tests, there are no requirements for PSL 1 Welded Pipe, so the requirements for PSL 2 Welded Pipe will be referenced instead. This edition of the specification th will be referred to as API 5L, 44 Edition throughout the report and accompanying tables. 2.4 Items Received for Testing On April 16, 2013 at approximately 1:50 pm CST, HurstLab received two (2) cut sections of pipe, and various other items from the failure location in Mayflower, Arkansas, which had been transported on a flatbed trailer. The two (2) sections of pipe were each wrapped in protective plastic with the open ends of the pipe sealed, and with the entire surface covered with plastic padding to protect from damage during loading/unloading and transportation. A 55 gallon steel drum, containing the coating that was removed in the field where the pipe was sectioned transversely, as well as a small bag containing possible calcareous deposits, were also received. The two (2) sections of pipe are described below in the same manner they are referenced throughout the report. 1) 33' 11-1/2" Long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it failed in service in Mayflower, Arkansas.

8 Page 5 of 185 2) 19' 10" Long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it failed in service in Mayflower, Arkansas. The Chain of Custody documents for the sections of pipe, as well as the steel drum of coating material and the possible calcareous deposits as well as the photographs documenting the evidence in the as-received condition are presented in Appendix II of this report. 3.0 METALLURGICAL EXAMINATION, TESTING AND ANALYSIS 3.1 Visual and Macroscopic Observations A 49' 9-1/2" long section of the Pegasus Pipeline, which fractured over a length of 22' along the ERW seam and 3" into the base metal at Milepost in Mayflower, Arkansas, as shown in Photographs No. 1 through No. 3, was removed from the ground by sectioning through three (3) locations of the pipeline following removal of the coating at those areas on the O.D. surface. The pipeline was transversely sectioned 3' upstream from the north girth weld through the adjoining intact pipe, 33' 11-1/2" from the north cut end, and 1' downstream from the south girth weld through the adjoining intact pipe The sections of pipe were received at HurstLab on April 16, The protective plastic, wrapping, and end plugs from both 33' 11-1/2" and 19' 10" long sections of the pipeline were carefully removed following receipt for examination and documentation of the evidence in the as-received condition, and to allow examination of the general condition of the pipe sections, such as the fracture, ERW seam and girth weld conditions, coating condition, evidence of any corrosion, mechanical damage, etc. Photographs No. 4 through No. 7 display the pipe sections in the as-received condition, and following removal of the plastic and wrapping. Examination of the 33' 11-1/2" long section of the pipe revealed a 22' long fracture along the ERW weld seam, which traversed diagonally, approximately 3" in length, into the base metal near the south end of the

9 Page 6 of 185 fracture. The fracture faces had been coated with a protective white grease in the field following the pipeline rupture to help preserve the fracture faces for subsequent analysis. All four (4) cut ends of the pipe sections were marked in the field denoting the location of the ERW seam, the relative position in ground, direction of the crude oil flow, station number and field cut match line in each section of the pipe. Photographs No. 8 and No. 9 display the as-received condition of the pipe and field markings on the pipe sections. 3.2 As-Received Condition of the Pipe and Coating Following unloading of the pipe from the transport truck and unwrapping of the protective material, the pipe was closely inspected to ascertain and document the as-received condition of the pipe and the coating. The 33' 11-1/2" long section of pipe contained a circumferential girth weld at the north end, and an approximately 3' long section of the adjoining intact pipe. The fracture, which followed the ERW seam at the 12:00 o clock position of the pipe, extended 22' 3" in length, with one fracture tip terminating in the north girth weld and the other in the base metal adjacent to the ERW seam. The maximum separation of the open crack was approximately 1-3/8" wide near the center of the crack, 12' from the north girth weld Examination of the coating showed a number of areas where the coating was damaged or split adjacent to the ERW seam. The maximum width and depth of the various splits in the coating on the O.D. surface of the pipe adjacent to the ERW seam, between the 10:30 and 1:30 o clock positions, were measured and photographically documented. Photographs No. 10 through No. 23 show the condition of the coating from 3' north of the north girth weld, referenced to as -3' from the north girth weld, to the girth weld at 0', and all the way to 50' 9-1/2" south of the north girth weld. As previously mentioned, the coating had been removed in the field from the areas where the pipe had been transversely sectioned.

10 Page 7 of 185 Coating Split Distance from North Girth Weld Maximum Width Maximum Depth Notes -3' 0' 1" * Some coating had been removed during sectioning in the field 0' 4' 2" 0.10" 4' 8' 0.5" 0.14" 8' 12' 0.5" * 12' 16' * 0.07" 16' 20' 0.25" 0.09" Longitudinal fracture or rupture of the pipe extended from the north girth weld at 0' to 22' 20' 24' 0.5" 0.10" 24' 28' 1.5" 0.10" 28' 30' 11-1/2" 1" 0.05" Some coating had been removed 30' 11-1/2" 35' 1" 0.15" during sectioning in the field 35' 39' 1" 0.10" 39' 43' 0.75" 0.11" 43' 47' 0.5" 0.11" 47' 50' 9-1/2" 1" * *Not measurable at location. Some coating had been removed during sectioning in the field The total thickness of the coating was estimated to be approximately 0.15" based on relatively intact areas of the coating, so some of the splits in the coating noted in the table above had likely penetrated to the base metal of the pipe. In addition to the splits noted above, the coating at the bottom, or 6 o clock position of the pipe was wrinkled, with the coating appearing to have sagged downward during the years the pipe lay buried. Although the coating did not appear stretched over the top and sides of the pipe, excess coating was folded over at the bottom of the pipe. Several places had small areas of coating missing, although it is not known at what point the coating loss had occurred during service. Additional photographs of the pipe and coating in the as-received condition are displayed in Photographs No. 24 through No. 64.

11 Page 8 of Coating Removal Process A procedure for a safe removal of the coating from the O.D. surface of the pipe was developed and approved by EMPCo and PHSMA, and is listed in Section A4 of the Test Protocol in Appendix I. The coating on the O.D. surface of the pipe was carefully removed on April 22, 2013 by Watkins Construction Company, LLC. (Watkins), a vendor contracted directly with EMPCo. Prior to proceeding, the contracted workers were briefed by HurstLab personnel as to the importance of preserving the fracture surface and integrity of the pipe; HurstLab personnel supervised the removal of the coating to ensure the safe removal of the coating. The coating on both pipe sections was first wet down with water, and each pipe section was then tightly wrapped in plastic wrap to securely collect all the coating. To remove the coating it was first cracked by tapping, and was then gently peeled off. First striking the coating with a resin hammer was tried; when the resin hammer did not crack the coating a steel mallet was used. The steel mallet was tapped against the coating, cracking the coating but not damaging the pipe underneath. The pipe sections were then cleaned using mineral spirits. Extreme care was taken to prevent any damage to the pipe or the fracture surface that could have affected the metallurgical investigation. All of the coating removed from the pipe sections at HurstLab, as well as the steel drum containing the coating that was removed in the field by EMPCo personnel, was collected and retained at EMPCo s facility in Corsicana, Texas. Appendix III shows several representative photographs of the coating removal process and contains the document signed by the employees of Watkins who removed the coating following the briefing by HurstLab personnel. 3.4 Condition of the Pipe Following Coating Removal Following removal of the O.D. coating in accordance with the specified guidelines, the pipe sections were re-examined to ascertain and photographically document the conditions of the pipe. The bottom of the

12 Page 9 of 185 pipe sections between approximately 4 and 8 o clock, at the locations where the coating had wrinkled and sagged, was covered with a reddishorange substance, likely a mixture of the surrounding native sandy soil that the pipe had been buried in and various corrosion products resulting from contact between the pipeline and moisture. Some corrosion pitting was visible within this area, as well as at various locations along the O.D. surface where the coating had previously split and allowed moisture to contact the surface of the pipe. No preferential or knife-like corrosion was present along the ERW seam at 12 o clock The depth of the corrosion pitting at the various locations around the O.D. surface of the fractured pipe section was measured using a certified and calibrated caliper, and the results are summarized in the following table. Distance from North Girth Weld Circumferential Location (o clock position) Depth of Corrosion Pitting Minimum Average Maximum -3' to 0' All No Corrosion Pitting Visible 0' to 4' 7:30 to 10: " 0.017" 0.029" 4' to 8' 8' to 12' 12' to 16' 16' to 20' 20' to 24' 1:30 to 3: " 0.013" 0.026" 6:45 to 10: " 0.013" 0.037" 3:45 to 5: " 0.011" 0.022" 7:30 to 11: " 0.011" 0.026" 3:00 to 5: " 0.013" 0.033" 6:30 to 10: " 0.017" 0.031" 2:45 to 5: " 0.015" 0.031" 7:00 to 10: " 0.012" :45 to 5: " 0.020" :15 to 10: " 0.010" ' to 28' All No Corrosion Pitting Visible 28' to 31' All No Corrosion Pitting Visible As shown, all of the corrosion pitting occurred between the 1:30 and 11:15 o clock positions on the fractured section of pipeline; no pitting corrosion was observed at the 12 o clock position where the ERW seam was positioned in the pipe. The average pitting depth over the entire section of the pipe was determined to be 0.014", and the maximum depth at any location was 0.037", which are approximately 4.5% and 12%, respectively, of the total wall thickness of the pipe. No corrosion pitting

13 Page 10 of 185 was present at either cut end of the fractured pipe section. Photographs showing the corrosion pitting on the east and west sides of the pipe following removal of the coating are displayed in Photographs No. 65 through No The I.D. surface of both pipe sections was examined using oblique lighting and pivoting mirrors and magnifying glasses prior to sectioning. No corrosion pitting was visible on the I.D. surface of either the fractured or intact sections of pipe. However some shallow bottomed depressions were observed at random locations. Following sectioning of the 33' 11-1/2" long and the 19' 10" long pipe lengths, the I.D. surfaces at several areas were more closely examined. Multiple shallow depressions, including those noted above, were visible around the entire circumference of the I.D. surface. The depressions were very smooth in appearance and contained no visible corrosion products, suggestive of mechanical deformation as opposed to corrosion pitting. No evidence of any significant corrosion pits was visible on the I.D. surface. Photographs No. 83 and No. 84 show representative areas of the I.D. surface. 3.5 Dimensional Measurements The out-of-roundness at intact locations at either end of the fracture, as well as at the south cut end of the 33' 11-1/2" long fractured section of pipe, was determined as specified in Section of API 5L, 44 th Edition. At each of the three (3) locations, four (4) measurements of the I.D. were taken, spanning between 12:00 and 6:00 o clock, 1:30 and 7:30 o clock, 3:00 and 9:00 o clock, and 4:30 and 10:30 o clock using a certified and calibrated I.D. micrometer. In accordance with the method th specified in the aforementioned section of API 5L, 44 Edition, the out-of- roundness at each location was then determined to be the difference between the largest and smallest I.D. measurement. The calculated outof-roundness at each location is displayed in the following table, along with the API requirements.

14 Page 11 of 185 Circumferential Location of Measurement (o clock) I.D. Measurement Distance from North Girth Weld Begins Ends -6" 271" 371" 12:00 6: " " " 1:30 7: " " " 3:00 9: " " " 4:30 10: " " " Calculated Out-of-Roundness 0.111" 0.036" 0.100" th API 5L, 44 Edition, Table 10, Pipe Except End Out-of-Roundness tolerance for D = 20" 0.400" As shown, at each of the locations tested the calculated out-of-roundness was determined to be within the allowable tolerance specified in API 5L, th 44 Edition, Table 10, for welded pipe with a nominal O.D. between 6.625" and 24". The results of the multiple I.D. measurements and the out-of-roundness calculations are recorded in Table Wall thickness measurements of the failed pipe were made at 2" intervals along the fracture adjacent to each mating fracture surface, using a certified and calibrated micrometer. The measurements were taken beginning at a location 40" south of the north girth weld and terminating at the crack tip, located 267", or 22' 3", from the north girth weld. Although the other crack tip was located at the north girth weld, the distance between the mating fracture surfaces was too small to allow for accurate wall thickness measurements at or directly adjacent to the north girth weld. The smallest wall thickness was measured to be 0.310" and the largest was 0.321". The average wall thickness was calculated to be 0.315", while the nominal specified wall thickness for the 20" O.D. pipe was 0.312". The complete results of the wall thickness measurements taken on either side of the crack using a certified and calibrated digital micrometer are recorded in Table 2.

15 Page 12 of The wall thickness of the fractured pipe was measured at numerous locations both at and away from the fracture by SGS-PfiNDE, Inc. (PfiNDE), an approved third party vendor using the non-destructive ultrasonic test method A grid or map of ultrasonic wall thickness measurements, covering from 12" upstream to 12" downstream of the fracture and around the entire 360 circumference of the pipe, were taken at 2" intervals over a total pipe length of 24.67'. The wall thickness was determined to range between 0.288" and 0.316" along the evaluated length. No internal corrosion areas were noted, although a linear inclusion in the mid-wall area of the pipe was noted on the CMAPPs (AUT) inspection. The complete results of the ultrasonic wall thickness measurements of the fractured pipe are recorded in Appendix IV. 3.6 Residual Stresses As the pipe containing the fracture was sectioned for fractographic examination, a significant amount of displacement of the sectioned portion of pipe was observed near the crack tip adjacent to the north girth weld, as shown in Photograph No. 85, indicating that the pipe had been under a considerable amount of constraint since it was manufactured, placing the ERW seam under sustained tension forces, which contributed to the increase in stresses at the ERW seam joint. The separation of the fracture faces confirms elastic spring back in the circumferential direction, indicating the presence of circumferential residue stresses likely associated with the original forming and ERW seam welding of the pipe. However, the extent to which these residual stresses may have contributed to the initiation of the hook cracks or the final fracture is unknown at this time. 3.7 Fractographic Examination The mating fracture faces of the entire 22' 3" long fracture were visually examined using oblique lighting prior to removal of the coal-tar coating, but following removal of the protective grease with mineral spirits, acetone, and a nylon brush. A thorough, careful examination of both

16 Page 13 of 185 mating fracture faces revealed fine chevrons or radial lines emanating from the fracture zone at a distance between 19' 10" and 21' 6-1/4" from the north girth weld, indicating that the final fracture, which resulted in the leakage of the crude oil, originated from this zone. Visual examination of the mating fracture faces from the distance between 1/4" and 26" south of the north girth weld revealed evidence of upturned grain flow lines or bands, and/or inclusions near the outer wall. However, there was no evidence of any chevron marks pointing to this fracture zone, indicating that the fracture did not initiate from this zone, but rather propagated through the surface imperfections. Photograph No. 86 displays overall and close-up views of the fracture origin and the tip areas, as well as field markings on the pipe. The fracture zones from a distance between 19' 10" and 20' from the north girth weld was further examined to characterize the fracture morphologies. Fractographic examination revealed flat, highly oxidized, fracture zones predominantly in the upper half (adjacent to the O.D. surface) of the fracture surface along the ERW seam, which are characteristic of hook cracks. Examination further revealed radial lines emanating from the tips of the hook cracks, indicating that the final fracture, which occurred during service and resulted in the leakage of the crude oil, originated from the tips of hook cracks that had reduced the effective cross-sectional area of the wall at the ERW seam location. A hook crack is defined in API Bulletin 5TL as Metal separations resulting from imperfections at the edge of the plate of skelp, parallel to the surface, which turn toward the inside diameter or outside diameter pipe surface when edges are upset during welding. Photograph No. 87 displays the final fracture initiation sites with insert photographs, revealing the hook cracks, final fracture zones, and the direction of the fracture propagation. The secondary fracture zone, found from a distance between 1/4" and 26" from the north girth weld, contained ERW seam manufacturing imperfections in the upset/haz area that had most likely cracked during the final rupture, and is displayed in Photographs No. 88 through No A section of the pipe containing the hook cracks, which measured approximately 3-1/2" to 4" in width and approximately 40" in length, was cut and removed from the pipe for closer examination of the O.D. and I.D.

17 Page 14 of 185 surfaces, and characterization of the fracture morphology. Photographs No. 95 and No. 96 display the cut sections. Close-up examination of the fracture face from a distance between 18' 10" and 19' 10-1/4" from the north girth weld revealed fine chevrons pointing to the hook cracks, indicating that the final fracture originated from the hook cracks and rapidly propagated upstream toward the north girth weld through the HAZ of the ERW seam. Photographs No. 97 through No. 100 display the evidence of chevrons pointing to the hook cracks. Further examination of the fracture face from a distance between 19' 10" and 20' 8" from the north girth weld revealed continuation of the hook cracks and transitioning of the radial lines into vertical lines, indicating the primary fracture origins to be between 20' 2-3/8" and 20' 7-3/8", as displayed in Photographs No. 101 through No Examination of the remaining fracture surface of the selected fracture face revealed continuation of the hook cracks with intermittent termination and continuation up to a location approximately 20' 11" from the north girth weld, and occasional hook cracks near the I.D. surface of the pipe with chevrons pointing in the opposite direction, indicating that the remaining final fracture propagated toward the south end and terminated in the base metal, as displayed in Photographs No. 104 through No In addition to the total depth of the hook cracks, the length and depth below the O.D. surface of various fracture zones on the fracture surface were measured as per the client s request. The darker smooth areas on the fracture surface, all beginning at the O.D. surface, indicated areas of the hook cracks that contained a tightly adhered layer of oxide scale from exposure to moisture; the length and maximum depth of each of these areas was measured. Several axial ridges were also visible on the fracture surface within the hook cracks, formed most likely as a result of the microstructural conditions of the upturned banded grain structure within the ERW seam upset and primary HAZ and potential microcracks through which the fracture occurred. The following table records the measurements, along with the distance from the north girth weld and reference to the photographs showing the various fractographic features.

18 Page 15 of 185 Fracture Zone Number Photograph Number Distance from North Girth Weld ' 3/8" to 20' 7/8" ' 2-1/8" to 20' 2-5/8" ' 3" to 20' 4-3/8" Feature Appearance Darker Smooth Area Darker Smooth Area Darker Smooth Area Total Length Depth Below O.D. Surface 1/2" 0.125" 1/2" 0.063" 1-3/8" 0.085" ' 3" to 20' 3-3/4" Ridge 3/4" 0.061" ' 3-7/8" to 20' 4-1/8" Ridge 1/4" 0.058" ' 4-5/8" to 20' 7-5/8" Darker Smooth Area 3" 0.150" ' 4-5/8" to 20' 6-3/8" Ridge 1-3/4" 0.113" ' 7-7/8" to 20' 8-1/8" ' 8-5/8" to 20' 9" ' 9-1/8" to 20' 11-1/4" ' 1/8" to 21' 1-1/2" ' 3" to 21' 4-3/8" ' 5" to 21' 5-1/2" ' 5-1/2" to 21' 5-7/8" Darker Smooth Area Darker Smooth Area Darker Smooth Area Darker Rough Area Darker Rough Area Darker Rough Area Darker Smooth Area 1/4" 0.046" 3/8" 0.063" 2-1/8" 0.048" 1-3/8" 0.062" 1-3/8" 0.031" 1/2" 0.042" 3/8" 0.020" An approximately 5-1/2" long section of the fracture surface containing the primary final fracture origins and some of the hook cracks between a distance of 20' 2-1/2" and 20' 8" from the north girth weld was removed, electrolytically descaled, cleaned using alkaline Endox 214 solution, and examined at low magnifications to ascertain the general condition of the pipe surface at the O.D. and I.D. surfaces along the ERW seam near the fracture origins. The mating fractured surface was not cleaned to preserve the sample for the later evaluation of the condition of the scale or oxidation that was present on the fractured face.

19 Page 16 of 185 Close-up examination of the cleaned fracture face containing hook cracks and the final fracture origins revealed that one of the final fracture origins was at a location where the outer coal-tar coating had split diagonally during service. Some of the coal-tar had melted onto the fracture surface. The examination also revealed localized melting of the pipe metal caused by the copper electrode contacts that were apparently originally used to weld the skelp to form the ERW pipe. Photographs No. 111 through No. 116 display the O.D. surface condition of the pipe near the fracture origins. Close-up examination of the fracture face between a distance of 20' 2-1/2" and 20' 8" from the north girth weld revealed highly oxidized hook cracks and the final fracture originating from the hook cracks, which were present to a maximum depth of 0.150". Photographs No. 117 through No. 122 display the hook cracks and the origins from where the final fracture initiated and propagated north toward the north girth weld along the ERW seam and south into the base metal south of the fracture origins The hook cracks and the final fracture zones across the entire fracture face from the O.D. to the I.D. of the pipe at two (2) of the several fracture origins, located at 20' 5-5/16" and 20' 6-3/4" from the north girth weld, as shown in Photographs No. 117 through No. 122, were examined at higher magnifications using a Scanning Electron Microscope (SEM) to further characterize the fracture morphologies. The SEM examination of the hook cracks revealed fractures through the multiple planes across the weld upset, HAZ, and/or fusion line of the ERW seam, which were covered with tightly adhered scale or oxidation products obscuring the fracture morphology. However, the fractures through multiple planes in the weld upset, HAZ, and/or fusion line suggest that the cracks propagated through the path of least resistance. There was some evidence of what appeared to be intergranular fracture in an extremely small area of the hook crack, which can be attributed to the prior grain structure of the material. The final fracture zone revealed essentially cleavage to quasicleavage fracture, indicative of brittle instantaneous failure. The fracture through the weld flash near the I.D. surface revealed evidence of ductile fracture. Photographs No. 123 through No. 150 document the fracture morphologies at the fracture origin locations.

20 Page 17 of Crack Measurements Fractographic examination of the fracture face between 19' 10" and 22' revealed the presence of the hook cracks along the multiple planes of the ERW seam between a distance of 19' 10-1/8" and 21' 9-1/2"; however, the hook cracks were predominantly located between 19' 10-1/8" and 20' 11-3/8", and 21' 2" and 21' 9-1/4", as measured from the north girth weld. The maximum depth of the hook cracks, from where the final fracture initiated during service and lead to the rupture of the pipeline, was 0.150"; however, the depth of the hook cracks varied between 0.016" and 0.150", as recorded in Table The mating fracture faces in the crack origins area from where the final fracture had initiated between a distance of 20' 2-1/2" and 20' 8" were reconstructed and sectioned transversely across the fractured ERW seam, more specifically at distances of 20' 3-3/4", 20' 4-7/8", and 20' 5-1/2" from the north of the girth weld, and were prepared for metallographic examination as well as the crack width measurements. Additional crosssections were also removed through the fractured ERW seam from a distance of 20' 6-13/16" and intact seam from a distance of 35' 8-1/2" and prepared for metallographic examination The maximum width and depth of the hook cracks were measured at several locations and were found to be " and 0.150", respectively. It should be noted here that the hook crack width measurements were made following reconstruction of the two (2) mating fracture faces and, therefore, the values shall be considered as approximates only. Table 4 records the hook cracks width measurements. 3.9 Metallographic Evaluation Microstructural examination of the cross-sections removed transversely through the ERW seam at a distance of 20' 4-7/8" and 20' 6-13/16" from the north girth weld and prepared for metallographic examination was performed to characterize the microstructural conditions of the ERW seam at the fracture origin locations. Microstructural examination revealed hook cracks through the ERW upset/haz along the realigned inclusions and upturned bands of extremely brittle untempered martensite.

21 Page 18 of 185 Both cross-sections removed through the final fracture origins and prepared for metallographic examination confirmed the presence of hook cracks through the excessive amount of manganese sulfide inclusions and bands which were essentially parallel to the ERW fusion line, an undesirable condition that was apparently created during the skelp forming and ERW processes. The microstructure of the upturned bands consisted of very brittle, hard untempered martensite, while the ERW upset/haz area consisted of a mixed-microstructure with grain boundary ferrite, unresolved bainite, and some untempered martensite, which is undesirable since this microstructure possesses extremely low ductility. The secondary HAZ and the base metal consisted of grain boundary ferrite and pearlite. Microstructural examination also revealed evidence of localized melting and cracking to a shallow depth at the electrode contact areas at the O.D. locations parallel to the weld seam. Photographs No. 151 through No. 202 document the microstructural condition of the ERW seam at the locations of the hook cracks from where the final fracture had initiated and predominantly propagated upstream toward the north girth weld A cross-section was removed transversely through the intact portion of the ERW seam of the 49' 9-1/2" section of the pipeline at a distance of 35' 8-1/2" from the north girth weld and prepared for metallographic examination to characterize the microstructural condition of the ERW seam. The microstructural examination revealed excessive amounts of predominantly manganese sulfide stringers and some oxide inclusions, several of them aligned parallel to the fusion line in the upset area of the ERW seam, which is a highly undesirable condition and can lead to the formation of hook cracks. The microstructural examination of the crosssection following etching in a 2% Nital solution revealed the presence of some upturned bands, however not as severe as those found in the fractured seam. The microstructure of the upturned bands consisted of brittle untempered martensite, while the upset/haz away from the bands consisted of mix-microstructure of grain boundary ferrite, bainite, and some untempered martensite. Photographs No. 203 through No. 220 document the microstructural condition of the intact ERW seam.

22 Page 19 of Longitudinal cross-sections were removed through the corrosion pitting at representative areas on the O.D. surface and through the shallow indentations on the I.D. surface, and were metallographically prepared and etched in a solution of 2% Nital. On the O.D. surface multiple pits filled with oxides and corrosion products were visible, extending to a maximum depth of 0.008" on the metallographically prepared crosssections. Following etching, the non-uniform pits were confirmed to be the result of material loss due to corrosion, with no evidence of grain deformation or mechanical damage. As previously noted, all of the corrosion pitting was observed between the 1:30 and 11:15 o clock positions on the fractured section of pipe, and no pitting corrosion was observed at the 12:00 o clock position where the ERW seam was positioned in the pipe. The corrosion observed on the O.D. surface did not contribute to the pipeline failure. Examination of the I.D. surfaces on the metallographically prepared cross-sections revealed that the shallow depressions were smooth indentations, between 0.137" and 0.189" wide and up to 0.007" deep. The I.D. surface and the surfaces of the indentations were smooth, with no visible oxide scale, and in the etched condition some grain deformation was visible at the edges of the indentations, indicating mechanical damage. However, the thickness of the microstructural band containing partial decarburization on the I.D. surface remained constant, indicating that the impressions occurred most likely during the hot-rolling of the steel or manufacturing of the pipe and not during service. The I.D. surface indentations did not contribute to the pipeline failure. Photographs No. 221 through No. 226 display representative areas of the O.D. and I.D. surfaces on the metallographically prepared longitudinal cross-sections in both the as-polished condition and following etching in a solution of 2% Nital Microhardness Surveys Vickers microhardness surveys were performed on the metallographically prepared cross-sections at both the representative fractured and intact locations of the ERW seam on the pipe sections in accordance with the 1 test method specified in ASTM E The Vickers microhardness

23 Page 20 of 185 values were converted to equivalent Rockwell B or C scale values based on the conversions provided in ASTM E140-07, Tables 1 and 2. It should be emphasized that the hardness equivalents are approximates based on equations developed from empirical data, and are typically higher than the results obtained by testing using the larger Rockwell indenter and much higher load forces Vickers microhardness surveys were performed on the metallographically prepared cross-sections removed from representative fractured areas of the ERW seam at 20' 4-7/8" and 20' 6-13/16" from the north girth weld. Each cross-section was evaluated along the fracture surface, including along the hook crack(s), the hardened martensitic upturned grains, and the final fracture zone, as well as in the ERW seam at the fusion line, the HAZ and the base metal. The results of the Knoop microhardness surveys at fractured locations of the pipe are summarized in the following table. Cross-section Location (from North Girth Weld) Base Metal Average Hardness, Rockwell Equivalent Heat- Affected Zone Hook Crack At Fracture Surface Hardened Upturned Grains Final Fracture ERW Fusion Line 20' 4-7/8" 96 HRB 100 HRB 29 HRC 52 HRC 28 HRC 42 HRC 20' 6-13/16" 100 HRB 21 HRC 29 HRC 49 HRC 29 HRC 32 HRC As shown, the hardness varied extensively along the fracture surface of the hook crack(s) within the upturned grains. The hardened, martensitic microstructure was 20 to 23 Rockwell C hardness points higher than the adjacent microstructure within the upturned grains and along the fusion line in the ERW seam. The hardness decreased the farther away from the ERW seam, resulting in approximately a 30 Rockwell C hardness point difference between the ERW seam and the softer base metal. The large difference in hardness is undesirable and results in increased internal stresses, which can contribute to crack initiation and propagation. The complete results of the Vickers microhardness surveys, including micrographs showing the locations of each indentation on the metallographically prepared cross-sections removed through the crack are displayed in Table 5 and Table 6.

24 Page 21 of A Vickers microhardness survey was also performed on the metallographically prepared cross-section that was removed through the ERW seam at a representative intact area approximately 35' 8-1/2" from the north girth weld for comparison with the data from the fractured location. The results of the Vickers microhardness survey of the intact area are displayed in the following table. Cross-section Location (from North Girth Weld 35' 8-1/2" Base Metal 100 HRB average Heat-Affected Zone 99 HRB average Hardness, Rockwell Equivalent Upturned Grain Flow Lines Varied between 21 HRC and 54 HRC ERW Fusion Line Varied between 23 HRC and 54 HRC As shown, the cross-section removed from an intact area of the pipe also contained a hardened martensitic microstructure within the upturned grain flow pattern of the ERW seam at the O.D. surface. The fusion line, HAZ, and base metal hardnesses of the intact cross-section were similar to those areas on the fractured cross-sections, including the large variation between the ERW seam and the base metal of the pipe. The complete results of the Vickers microhardness survey, including a micrograph of the metallographically prepared cross-section removed from the ERW seam in an intact area, are displayed in Table Tensile Tests In order to determine the ultimate tensile stress, yield stress at a 0.5% offset, and percent elongation of the pipe, multiple tensile test specimen blanks were removed through the ERW seam, as well as in both the transverse and longitudinal directions away from the seam, on the intact 19' 10" long section of pipe as shown in Appendix V. All of the test specimens were machined to have a 2" long gauge length, a 1-1/2" wide reduced section, and represented essentially the entire wall thickness, with only slight sanding to remove minor surface imperfections or, as noted, the weld flash Six (6) transverse tensile test specimen blanks were removed through the th ERW seam and were then flattened as specified in both the 10 Edition th and the 44 Edition of API 5L. The tensile test specimens were then

25 Page 22 of 185 machined and tested in accordance with ASTM A370-12a and the applicable sections of each edition of the API 5L specification. The results of the transverse tensile tests through the ERW seam, along with the th tensile requirements from both the 10 Edition of API 5-L that was in effect at the time the pipe was manufactured and the current API 5L, 44 th Edition are shown in the following table. Sample Ultimate Yield Fracture Identification Stress (psi) Stress (psi) Elongation (%) Location Transverse, Through ERW Seam, Weld Flash Included, Sample 1 Transverse, Through ERW Seam, Weld Flash Included, Sample 2 Transverse, Through ERW Seam, Weld Flash Included, Sample 3 Transverse, Through ERW Seam, Weld Flash Removed, Sample 1 Transverse, Through ERW Seam, Weld Flash Removed, Sample 2 Transverse, Through ERW Seam, Weld Flash Removed, Sample 3 101,000 77,000 4 HAZ 93,500 79,000 5 HAZ 102,000 84, Base Metal 85,500 73,000 3 HAZ 85,500 75,000 3 HAZ 92,500 77,000 5 HAZ th API 5-L, 10 Edition, Electric Welded Pipe, Open Hearth Steel, Grade B 60,000 minimum None Specified None Specified Not Applicable th API 5L, 44 Edition, PSL 1, 60,200 None None Not Welded Pipe, Grade X42 minimum Specified Specified Applicable As shown, all of the tensile test specimens, regardless of whether the specimens contained the weld flash, met the minimum ultimate stress th th requirements specified in both API 5-L, 10 Edition and API 5L, 44 Edition. The complete results of the transverse tensile tests through the ERW seam are recorded in Table Multiple base metal transverse tensile test specimen blanks were removed o from the pipe, at locations 90 from the ERW seam and 180 from the ERW seam, and were flattened prior to machining. Longitudinal base metal tensile test specimen blanks were also removed from the pipe at a o location 90 from the ERW seam. All of the tensile test blanks were

26 Page 23 of 185 machined and tested in accordance with ASTM A370-12a and the applicable sections of sections of each edition of API 5L. The results of both the transverse and longitudinal base metal tensile tests, along with th the tensile requirements from both the 10 Edition of API 5-L that was in effect at the time the pipe was manufactured and the current API 5L, 44 th Edition are shown in the following table. Sample Identification Transverse, 90 from ERW Seam, Sample 1 Transverse, 90 from ERW Seam, Sample 2 Transverse, 90 from ERW Seam, Sample 3 Transverse, 180 from ERW Seam, Sample 1 Transverse, 180 from ERW Seam, Sample 2 Transverse, 180 from ERW Seam, Sample 3 Longitudinal, 90 from ERW Seam, Sample 1 Longitudinal, 90 from ERW Seam, Sample 2 Longitudinal, 90 from ERW Seam, Sample 3 Ultimate Stress (psi) Yield Stress (psi) Elongation (%) 87,000 59, ,500 59, ,000 62, ,000 63, ,500 60, ,500 64, ,000 64, ,000 66, ,500 68, th API 5-L, 10 Edition, Electric Welded Pipe, Open Hearth Steel, Grade B th API 5L, 44 Edition, PSL1, Welded Pipe, Grade X42 60,000 minimum 60,200 minimum 35,000 minimum 42,100 minimum Unknown 1 27% minimum 1 The required minimum elongation specified on the tensile requirements table in the th provided paper copy of API 5-L, 10 Edition is illegible. As shown, all of the base metal tensile test specimens, in both the transverse and longitudinal directions, met the requirements specified in th th both API 5-L, 10 Edition and API 5L, 44 Edition. Although the measured yield stress typically exceeded the minimum ultimate stress requirement, it should be noted that there were not any maximum strength requirements. The complete results of the base metal transverse and longitudinal tensile tests are recorded in Tables 9 and 10.

27 Page 24 of Sub-sized round, non-flattened transverse tensile test specimen blanks were removed through the ERW seam, 90 from the ERW seam, and 180 from the ERW seam on the intact section of pipe, and were machined and tested in accordance with the applicable sections of API 5L and ASTM A370-12a. The results of the non-flattened transverse tensile tests are summarized in the following tables. Sample Identification Transverse, Through ERW Seam, Weld Flash Removed, Non-flattened Ultimate Yield Stress (psi) Stress (psi) Elongation (%) 99,600 65, th API 5-L, 10 Edition, Electric Welded 60,000 None None Pipe, Open Hearth Steel, Grade B minimum Specified Specified th API 5L, 44 Edition, PSL1, 60,200 None None Welded Pipe, Grade X42 minimum Specified Specified Sample Identification Transverse, 90 from ERW Seam, None-flattened Transverse, 180 from ERW Seam, None-flattened Ultimate Stress (psi) Yield Stress (psi) Elongation (%) 86,100 56, ,600 57, th API 5-L, 10 Edition, Electric Welded Pipe, Open Hearth Steel, Grade B 60,000 minimum 35,000 minimum Unknown 1 th API 5L, 44 Edition, PSL1, 60,200 42,100 27% Welded Pipe, Grade X42 minimum minimum minimum 1 The required minimum elongation specified on the tensile requirements table in the th provided paper copy of API 5-L, 10 Edition is illegible. As shown, the sub-sized, non-flattened transverse tensile test specimens th met the requirements specified in both API 5-L, 10 Edition and API 5L, th 44 Edition. The complete results of the sub-sized, non-flattened transverse tensile tests are recorded in Table Charpy V-Notch Impact Tests Test blanks for multiple sets of transverse Charpy V-Notch (CVN) impact test specimens were removed from the intact 19' 10" long section of pipe as shown in Appendix V. Sets of half-sized 10 mm x 5 mm test th specimens were machined per Section 9.8 of API 5L, 44 Edition and

28 Page 25 of 185 ASTM A370-12a and were notched in the fusion line of the ERW seam, the primary HAZ of the ERW approximately 1 mm from the fusion line, and the base metal. Then for each notch location, one (1) set of three (3) specimens was tested per ASTM A370-12a at the selected test temperatures of plus 32 F, plus 65 F, plus 80 F, and plus 95 F. Base metal specimens were also tested at additional temperatures The results of the CVN impact tests for each location and each test temperature are recorded in the following tables. V-Notch Location: ERW Fusion Line Specimen Number Test Temperature Impact Value (ft-lbf) Lateral Expansion (mils) Percent Shear (%) Plus 95 F Plus 80 F Plus 65 F Plus 32 F V-Notch Location: ERW Primary Heat-Affected Zone Specimen Number Test Temperature Impact Value (ft-lbf) Lateral Expansion (mils) Percent Shear (%) Plus 95 F Plus 80 F Plus 65 F Plus 32 F

29 Page 26 of 185 Specimen Number 1 Test Temperature V-Notch Location: Base Metal Impact Value (ft-lbf) Lateral Expansion (mils) Percent Shear (%) Plus 95 F Plus 80 F Plus 65 F Plus 32 F Zero F Minus 32 F As shown, the impact values at each notch location were essentially the same between plus 32 F and plus 95 F, while the base metal impact values at 0 F were half the values at 32 F and above, and continued to drop with lower temperatures. The fusion line of the ERW seam had the lowest impact values and the base metal, as expected, had the highest values. The lateral expansion and percent shear was essentially zero at the fusion line of the ERW seam, and the lateral expansion was only slightly higher in the HAZ. The base metal had the largest lateral expansion and percent shear values. The results of the CVN impact tests are recorded in Tables 12, 13, and 14. At the time the pipe was manufactured, no CVN impact tests or th requirements were specified in APL 5-L, 10 Edition. Likewise, there are no impact requirements for Type PSL 1 welded pipe in the current 44 th Edition of API 5L. The only impact requirements for comparison are that th in the 44 Edition of API 5L, for all notch locations on Type PSL 2 welded pipe, Grade X60, half-size transverse test specimens are required to have a 10 ft-lbf minimum average for a set of three test specimens and 8 ft-lbf minimum for a single individual test specimen, when tested at a test temperature of plus 32 F.

30 Page 27 of The CVN impact test results were then intended to be used to determine the lower shelf energy, upper shelf energy, the ductile-to-brittle transition temperature for the base metal, and if possible, the ERW seam, by plotting the results and developing an S-curve graph. The ductile-to-brittle transition temperature for the ERW fusion line and HAZ can not be determined, because the results of the impact tests at these areas were essentially the same regardless of test temperature. All of the CVN impact test specimens notched in the ERW seam, whether at the fusion line or in the HAZ, failed in an essentially brittle manner, therefore the ductile-to- o brittle transition temperature is above 95 F and is outside the scope of this investigation. However, additional tests at a temperatures below plus 32 F were performed on transverse CVN impact test specimens machined from the base metal because the base metal test specimens did fracture in a more ductile manner. The lower shelf would be considered to be around 2 ft-lbf for the size tested, or 4 ft-lbf for a full-size test specimen Chemical Analyses An approximately 2" by 2" section was removed away from the ERW seam on the intact 19' 10" long section of pipe, as shown in Appendix V, and the surface was sanded smooth in preparation for determining the chemical composition of the pipe using the Optical Emission Spectroscopic (OES) test method in accordance with ASTM E415-08, with the percent carbon determined by an approved vendor using the combustion method specified in ASTM E The results of the chemical composition analysis, as well as the compositional requirements for both the 10 th Edition of API 5-L that was in affect at the time the pipe was th manufactured and the current API 5L, 44 Edition are shown in the following table.

31 Page 28 of 185 Element (wt%) Sample Tested th API 5-L, 10 Edition, Electric Welded Pipe, Open Hearth Steel, Grade B Spec. th API 5L, 44 Edition, PSL 1, Welded Pipe, Grade X42 Specification Carbon max 0.26 max Manganese to max Phosphorus max max Sulfur max max Silicon <0.01 Chromium <0.01 Nickel 0.04 Molybdenum <0.01 Copper 0.02 Aluminum <0.01 Niobium <0.01 Vanadium <0.01 Titanium < max 0.50 max 0.15 max 0.50 max Base Base Base 1 Analytical range not specified for element. 2 Sum of Niobium + Vanadium + Tantalum = 0.15% maximum As shown, the pipe met the chemical composition that was specified in th API 5-L, 10 Edition at the time of the pipe manufacture, but does not meet the compositional requirements specified in the current API 5L, 44 th Edition for welded pipe. The complete results of the OES chemical analysis of the pipe are recorded in Table The foreign materials on the fracture surfaces, the O.D. surface, and the tightly adhered, very viscous black coating of the pipe was analyzed using the Energy Dispersive X-ray Spectroscopic (EDS) test method in accordance with ASTM E a in order to determine the elements present and the relative amounts of each. It should be noted that the fracture surface was protected with white grease prior to shipment to the laboratory, which was removed with the mineral spirit and acetone, and therefore the results of the EDS analysis may not be taken at the face value. Furthermore, it should also be noted that EDS is a semiquantitative test method, and that the results should be used as comparative or relative values only. It should also be noted that the EDS used was not capable of detecting light elements, those elements with atomic weights less than fluorine.

32 Page 29 of 185 The following table shows the results of the EDS analysis at three (3) different locations of the fracture surface. Element (wt%) Fracture Surface EDS-1 Fracture Surface EDS-2 Fracture Surface EDS-3 Magnesium Aluminum Silicon Sulfur Chlorine Potassium Calcium Titanium Manganese Iron Element not detected. 1 As shown, in addition to iron and manganese from the base metal of the pipe, high levels of silicon, aluminum, and magnesium, were detected, most likely due to soil adhering to the fracture surface; similarly the calcium, potassium, and titanium were also likely from the surrounding soil. High levels of the corrosive elements chlorine and sulfur were also detected, although no pitting corrosion had yet occurred on the fracture surfaces. The complete results of the EDS analyses of the material on the fracture surfaces, including line spectra and SEM images of each location, are recorded in Tables 16, 17, and The chemical composition of the reddish-brown products on the O.D. surface of the pipe was also evaluated using the EDS test method. The results of the EDS analysis are displayed in the following table. Reddish-Brown Element (wt%) Product on O.D. Magnesium Aluminum Silicon Sulfur Potassium Titanium Manganese Iron

33 Page 30 of 185 As shown, the products on the O.D. surface of the pipe were composed of primarily silicon with aluminum and potassium, in addition to the iron from the base metal of the pipe. The reddish-brown product on the O.D. surface of the pipe was likely soil that had migrated through the splits in the coating of the pipe. Some of the products may also have been from corrosion of the pipe, although it should be stressed that there was no evidence of significant localized or pitting corrosion on the received sections of pipe. The results of the EDS analysis of the products on the O.D. surface of the pipe are recorded in Table The viscous black bitumen, or coal-tar, coating that was on the O.D. surface of the pipe underneath the layer of fibrous coating was also analyzed using the EDS test method. The results of the test are displayed in the following table. Element (wt%) Black Bitumen Coating Magnesium Aluminum Silicon Sulfur Silver No specific chemical composition of the coating was available for comparison. Bitumen is a highly viscous mixture composed primarily of highly condensed polycyclic aromatic hydrocarbons that is used as a waterproof coating for buried pipe, among other uses such as paving roads. The results of the EDS analysis of the viscous black coating on the O.D. surface of the pipe are recorded in Table CONCLUSION 4.1 Technical Causes of Failure Based on the inspection, testing, and evaluation performed in accordance with the approved metallurgical test protocol, review of the background information, and technical research, the following is HurstLab s opinion.

34 Page 31 of 185 The failure of the pipeline at Milepost in the Conway to Corsicana section of the Pegasus crude oil pipeline located in Mayflower, Arkansas, which occurred at 2:37 pm CST on March 29, 2013, resulted because of the reduction of the wall thickness in the upset zone of the Electric Resistance Weld (ERW) seam caused by the presence of manufacturing defects, namely the upturned bands of brittle martensite, combined with localized stress concentrations at the tips of the hook cracks, low fracture toughness of the material in the upset/haz, excessive residual stresses in the pipe from the initial forming and seam and girth welding processes, and the internal pressure creating hoop stresses. The hook cracks, with maximum dimensions of " in width, 0.150" in depth, and 13-1/4" in length, as measured on the examined section of the fracture surface, were present in the ERW seam prior to the incident for an unknown period of time. The weak upturned fibers or bands of untempered brittle martensite were created during the manufacturing of the pipe. The presence of the tightly adhered scale or oxidation products on the fracture faces of the hook cracks suggests that the hook cracks had been present for an unknown period of time. It is unclear, however, whether the hook cracks occurred immediately after manufacturing or during service. The hook cracks initiated and followed the brittle upturned grain flow lines or bands that were created during the manufacturing of the pipe due to effects of the stresses induced by hydrostatic testing, thermal stresses, residual stresses, and/or pressure cycles. The hook cracks may not have all occurred simultaneously, as suggested by variation in coloration of the scale or oxides on the fracture surface and the macroscopic features of the fracture. The hook cracks and potential microcracks in the upset/heat-affected zones may have then merged due to stresses during service. 4.2 Failure Scenario Based on the preceding conclusion, the evidence of the hook cracks through multiple ductile and brittle zones, significant variance in hardness between the various zones of the ERW seam, the tightness and depth of the hook cracks along multiple planes through the upset

35 Page 32 of 185 heat-affected zones, and the extremely low impact toughness and elongation properties across the ERW seam, it is highly probable that some micro-cracking within the upset/heat-affected zones might have occurred immediately following the pipe manufacturing. The micro-cracks then likely would have merged by further cracking through the adjacent areas in the localized upset/haz zones during service, forming a continuous hook crack in each of the localized areas to the critical depths, at which point the remaining wall thickness, combined with the localized stress concentration and the residual stresses, could no longer support the internal hoop stresses and resulted in the final failure. Submitted by, Mahesh J. Madhani Chief Metallurgist Revised on July 9, 2013 to clarify the findings and to make editorial changes. W:\1H.M.R.L.\1REPORTS\EXXONMOBIL PIPELINE COMPANY.64961, REV. 1.RLC OUR LETTERS AND REPORTS APPLY ONLY TO THE SAMPLE TESTED AND/OR EVALUATED AND ARE NOT NECESSARILY INDICATIVE OF THE QUALITIES OF APPARENTLY IDENTICAL OR SIMILAR PRODUCTS. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT.

36 Page 33 of 185 MP Photograph No. 1 The photographs provided by EMPCo of the 20" O.D. x 0.312" wall pipe at Milepost of the Conway to Corsicana Pegasus crude oil pipeline, which failed on Friday, March 29, 2013 at 2:47 pm CST in Mayflower, Arkansas, display a straight, linear crack at approximately the 12:00 o clock position.

37 Page 34 of 185 Girth Weld Photograph No. 2 Crack terminating in base metal Photograph No. 3 The photographs display close-up views of the crack tips near the north girth weld in the ERW seam of the pipe and the south end in the base metal, respectively.

38 Page 35 of 185 Photograph No. 4 Photograph No. 5 The photographs display the fractured section of the pipe in the as-received condition and following removal of the outer protective wrapping material.

39 Page 36 of 185 Photograph No. 6 The photograph displays the intact section of the pipe in the as-received condition with the outer protective wrapping material.

40 Page 37 of 185 Photograph No. 7 The photograph displays the intact section of the pipe following removal of the nd 2 protective wrapping material.

41 Page 38 of 185 Photograph No. 8 The photograph displays the fractured pipe section following removal of the nd 2 wrapping material, revealing the fracture faces coated with grease to protect from post-incident corrosion.

42 Page 39 of 185 Photograph No. 9 The photograph displays the intact section of the pipe following removal of st the 1 protective wrapping material.

43 Page 40 of 185 Photograph No. 10 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 11 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum -3' to 0' 1" - 0' to 4' 2" 0.10" The photographs display overall top views of the pipe adjacent to the fractured pipe from approximately 3' north of the north girth weld (-3') to the center of the north girth weld (0'), and the fractured pipe from the center of the girth weld to 4' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The fracture in the pipe along the ERW seam terminated at the north girth weld. The fracture was extremely tight at the girth weld but was measured to be approximately 13/16" in width approximately 4' south of the north girth weld. Relatively narrow longitudinal and transverse splits were present in the coating. The coating had been removed from the adjacent intact pipe prior to sectioning approximately 3' north of the north girth weld.

44 Page 41 of 185 Photograph No. 12 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 13 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 4' to 8' 0.5" 0.14" 8' to 12' 0.5" - The photographs display overall top views of the fractured pipe from 4' south to 8' south of the north girth weld, and from 8' south to 12' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. Longitudinal and transverse splitting is present in the coating, and some of the coating is missing on either side of the fracture.

45 Page 42 of 185 Photograph No. 14 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 15 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 12' to 16' " 16' to 20' 0.25" 0.09" The photographs display overall top views of the fractured pipe from 12' south to 16' south of the north girth weld, and from 16' south to 20' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. Longitudinal and transverse splitting is present in the coating, and some of the coating is missing on either side of the fracture.

46 Page 43 of 185 Photograph No. 16 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 17 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 20' to 24' 0.5" 0.10" 24' to 28' 1.5" 0.10" The photographs display overall top views of the fractured pipe from 20' south to 24' south of the north girth weld, and from 24' south to 28' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. At approximately 22' south of the girth weld, the fracture in the ERW seam turned into the pipe material, progressing several inches prior to terminating. The damaged area of coating near the pipe fracture extended longitudinally past the fracture tip several feet.

47 Page 44 of 185 Photograph No. 18 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 19 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 28' to 31' 1" 0.05" 31' to 35' 1" 0.15" The photographs display overall top views of the fractured pipe from 28' south to 31' south of the north girth weld, and from 31' south to 35' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The approximately 49' 9-1/2" long pipe was sectioned in the field transversely approximately 31' south of the north girth weld. The coating was removed in the field approximately 13" in either direction from the transverse cut prior to sectioning.

48 Page 45 of 185 Photograph No. 20 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 21 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 35' to 39' 1" 0.10" 39' to 43' 0.75" 0.11" The photographs display overall top views of the fractured pipe from 35' south to 39' south of the north girth weld, and from 39' south to 43' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. Longitudinal splitting is visible on the surface of the coating.

49 Page 46 of 185 Photograph No. 22 Circumferential Location 10:30 o clock to 1:30 o clock Photograph No. 23 As-received Condition of the Coating Distance from North Girth Weld Split Width Maximum Split Depth Maximum 43' to 47' 0.5" 0.11" 47' to 51' 1" - The photographs display overall top views of the fractured pipe from 43' south to 47' south of the north girth weld, and from 47' south to 51' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. Longitudinal splitting is visible on the surface of the coating. Some of the coating had been removed from the adjacent area pipe prior to sectioning.

50 Page 47 of 185 Photograph No. 24 Photograph No. 25 The photographs display overall views of the west side of the pipe from 7:30 to 10:30 o clock, adjacent to the fractured pipe from approximately 3' north of the girth weld (-3') to the center of the north girth weld (0'), and the fractured pipe from the center of the girth weld to 4' south of the north girth weld (+4'), respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains disbonded and wrinkled coating.

51 Page 48 of 185 Photograph No. 26 The photograph displays an overall view of the west side from 7:30 to 10:30 of the fractured pipe from 4' south to 8' south of the north girth weld in the as-received condition prior to removing the coating. The lower half of the pipe contains disbonded and wrinkled coating, and some openings in the coating are present where the coating had begun to sag.

52 Page 49 of 185 Photograph No. 27 Photograph No. 28 The photographs display overall views of the west side between 7:30 and 10:30 of the fractured pipe from 12' south to 16' south of the north girth weld, and from 16' south to 20' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains disbonded and wrinkled coating, along with some openings in the coating.

53 Page 50 of 185 Photograph No. 29 Photograph No. 30 The photographs display overall views of the west side from 7:30 to 10:30 of the fractured pipe from 20' south to 24' south of the north girth weld, and from 24' south to 28' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains disbonded and wrinkled coating.

54 Page 51 of 185 Photograph No. 31 Photograph No. 32 The photographs display overall views of the west side from 7:30 to 10:30 of the fractured pipe from 28' south to 31' south of the north girth weld, and from 31' south to 35' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The pipe had been sectioned transversely approximately 31' south of the north girth weld. The lower half of the pipe contains disbonded and wrinkled coating.

55 Page 52 of 185 Photograph No. 33 Photograph No. 34 The photographs display overall views of the west side between 7:30 and 10:30 of the fractured pipe from 35' south to 39' south of the north girth weld, and from 39' south to 43' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains disbonded and wrinkled coating.

56 Page 53 of 185 Photograph No. 35 Photograph No. 36 The photographs display overall views of the west side from 7:30 to 10:30 of the fractured pipe and adjacent intact pipe from 43' south to 47' south of the north girth weld, and from 47' south to 51' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating had been removed from the adjacent intact pipe prior to allow for sectioning. The lower half of the pipe contains disbonded and wrinkled coating.

57 Page 54 of 185 Photograph No. 37 Photograph No. 38 The photographs display overall bottom views of the pipe from 4:30 to 7:30 o clock adjacent to the fractured pipe from approximately 3' north of the north girth weld (-3') to the center of the north girth weld (0'), and the fractured pipe from the center of the north girth weld to 4' south of the north girth weld (+4'), respectively, in the as-received condition prior to removing the coating. The coating had been removed from the adjacent intact pipe prior to sectioning in the field. The coating on the lower half of the pipe is sagging and contains wrinkles.

58 Page 55 of 185 Photograph No. 39 Photograph No. 40 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 from 4' south to 8' south of the north girth weld, and from 8' south to 12' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe is sagging and contains wrinkles.

59 Page 56 of 185 Photograph No. 41 Photograph No. 42 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 from 12' south to 16' south of the north girth weld, and from 16' south to 20' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe contains wrinkles and has sagged.

60 Page 57 of 185 Photograph No. 43 Photograph No. 44 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 from 20' south to 24' south of the north girth weld, and from 24' to 28' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe contains a significant amount of wrinkles and has sagged quite a bit.

61 Page 58 of 185 Photograph No. 45 Photograph No. 46 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 from 28' south to 31' south of the north girth weld, and from 31' south to 35' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The fractured pipe was sectioned transversely approximately 31' south of the north girth weld into two sections.

62 Page 59 of 185 Photograph No. 47 Photograph No. 48 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 from 35' south to 39' south of the north girth weld, and from 39' south to 43' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe had sagged quite a bit and contains a significant amount of wrinkles.

63 Page 60 of 185 Photograph No. 49 Photograph No. 50 The photographs display overall bottom views of the fractured pipe from 4:30 to 7:30 and the adjacent intact pipe from 43' south to 47' south of the north girth weld, and from 47' south to 51' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe contains a significant amount of wrinkles. The coating on the lower half of the pipe contains wrinkles.

64 Page 61 of 185 Photograph No. 51 Photograph No. 52 The photographs display overall views of the east side of a pipe from 1:30 to 4:30 adjacent to the fractured pipe from approximately 3' north of the north girth weld (-3') to the center of the north girth weld (0'), and the fractured pipe from the center of the north girth weld to 4' south of the north girth weld (+4'), respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains wrinkled coating.

65 Page 62 of 185 Photograph No. 53 Photograph No. 54 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 from 4' south to 8' south of the north girth weld, and from 8' south to 12' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains sagging and wrinkled coating.

66 Page 63 of 185 Photograph No. 55 Photograph No. 56 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 from 12' south to 16' south of the north girth weld, and from 16' south to 20' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains sagging and wrinkled coating.

67 Page 64 of 185 Photograph No. 57 Photograph No. 58 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 from 20' south to 24' south of the north girth weld, and from 24' south to 28' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half of the pipe contains sagging and wrinkles.

68 Page 65 of 185 Photograph No. 59 Photograph No. 60 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 from 28' south to 31' south of the north girth weld, and from 31' south to 35' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The lower half of the pipe contains sagging and wrinkled coating.

69 Page 66 of 185 Photograph No. 61 Photograph No. 62 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 from 35' south to 39' south of the north girth weld, and from 39' south to 43' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating on the lower half is wrinkled and sagging.

70 Page 67 of 185 Photograph No. 63 Photograph No. 64 The photographs display overall views of the east side of the fractured pipe from 1:30 to 4:30 and adjacent intact pipe from 43' south to 47' south of the north girth weld, and from 47' south to 51' south of the north girth weld, respectively, in the as-received condition prior to removing the coating. The coating had been removed from the adjacent intact pipe prior to sectioning. The coating on the lower half of the pipe is wrinkled and sagging.

71 Page 68 of 185 Photograph No. 65 Circumferential Location Photograph No. 66 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum All -3' to 0' No Corrosion Pitting Visible 7:26 o clock to 10:07 o clock 0' to 4' 0.006" 0.017" 0.029" The photographs display overall views of the west side of the pipe adjacent to the fractured area of the pipe, from approximately 3' north of the north girth weld (-3') to the center of the girth weld (0'), and the fractured pipe from the center of the girth weld to 4' south of the girth weld (+4'), respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

72 Page 69 of 185 Photograph No. 67 Circumferential Location Photograph No. 68 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 6:41 o clock to 10:07 o clock 4' to 8' 0.002" 0.013" 0.037" 7:03 o clock to 11:16 o clock 8' to 12' 0.002" 0.011" 0.026" The photographs display overall views of the west side of the fractured pipe from 4' south to 8' south of the north girth weld, and 8' south to 12' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

73 Page 70 of 185 Photograph No. 69 Circumferential Location Photograph No. 70 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 6:29 o clock to 9:55 o clock 12' to 16' 0.003" 0.017" 0.031" 6:52 o clock to 10:07 o clock 16' to 20' 0.006" 0.012" 0.021" The photographs display overall views of the west side of the fractured pipe from 12' south to 16' south of the north girth weld, and 16' south to 20' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

74 Page 71 of 185 Photograph No. 71 Circumferential Location Photograph No. 72 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 7:15 o clock to 9:55 o clock 20' to 24' 0.005" 0.010" 0.021" All 24' to 28' No Corrosion Pitting Visible The photographs display overall views of the west side of the fractured pipe from 20' south to 24' south of the north girth weld, and 24' south to 28' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

75 Page 72 of 185 Circumferential Location Photograph No. 73 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum All 28' to 31' No Corrosion Pitting Visible The photograph displays an overall view of the west side of the fractured pipe from 28' south to 31' south of the north girth weld, respectively, after the removal of the coating. The fractured pipe was sectioned in the field transversely approximately 31' south of the north girth weld to allow for removal of the fractured section of pipe. No corrosion pitting is visible on the O.D. surface near the transverse cut at the south end of the fractured section of the pipe.

76 Page 73 of 185 Photograph No. 74 Circumferential Location Photograph No. 75 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum All -3' to 0' No Corrosion Pitting Visible All 0' to 4' No Corrosion Pitting Visible The photographs display overall views of the east side of the pipe adjacent to the fractured pipe from approximately 3' north of the girth weld (-3') to the center of the north girth weld (0'), and the fractured pipe from the center of the girth weld to 4' south of the north girth weld (+4'), respectively, after the removal of the coating. No corrosion pitting is visible on the O.D. surfaces on the fractured or intact pipe around the north girth weld.

77 Page 74 of 185 Photograph No. 76 Circumferential Location Photograph No. 77 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 1:31 o clock to 3:03 o clock 4' to 8' 0.008" 0.013" 0.026" 3:49 o clock to 4:57 o clock 8' to 12' 0.004" 0.011" 0.022" The photographs display overall views of the east side of the fractured pipe from 4' south to 8' south of the north girth weld, and 8' south to 12' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

78 Page 75 of 185 Photograph No. 78 Circumferential Location Photograph No. 79 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 3:03 o clock to 4:57 o clock 12' to 16' 0.003" 0.013" 0.033" 2:40 o clock to 5:20 o clock 16' to 20' 0.005" 0.015" 0.031" The photographs display overall views of the east side of the fractured pipe from 12' south to 16' south of the north girth weld, and 16' south to 20' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

79 Page 76 of 185 Photograph No. 80 Circumferential Location Photograph No. 81 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum 2:40 o clock to 4:57 o clock 20' to 24' 0.004" 0.020" 0.033" All 24' to 28' No Corrosion Pitting Visible The photographs display overall views of the east side of the fractured pipe from 20' south to 24' south of the north girth weld, and 24' south to 28' south of the north girth weld, respectively, after the removal of the coating. The lower half of the pipe shows corrosion pitting on the O.D. surface where the coating had wrinkled and sagged.

80 Page 77 of 185 Circumferential Location Photograph No. 82 Distance from North Girth Weld Depth of Corrosion Pitting Minimum Average Maximum All 28' to 31' No Corrosion Pitting Visible The photograph displays an overall view of the east side of the fractured pipe from 28' south of the north girth weld to 31' south of the north girth weld, respectively, after the removal of the coating. No corrosion pitting was visible on the O.D. surface near the transverse cut at the south end of the fractured section of the pipe.

81 Page 78 of 185 Photograph No. 83 Photograph No. 84 The photographs display representative areas of the I.D. surface at an intact area of the pipe, showing the smooth, shallow impressions that resulted from mechanical damage, most likely during the hot-rolling of the steel or manufacturing of the pipe. No evidence of corrosion pitting was observed on the I.D. surface.

82 Page 79 of 185 Photograph No. 85 The photograph shows the displacement of the pipe by approximately 2-31/32" following sectioning through the intact portion of the adjoining pipe, indicative of the presence of significant residual stress.

83 Page 80 of 185 Fracture Initiation Sites Photograph No. 86 The photographs display overall and close-up views of the 33' 11-1/2" long section of a fractured 20" O.D. x 0.312" wall pipe, which was removed from the Conway to Corsicana section of the Pegasus Crude Oil Pipeline at Milepost in Mayflower, Arkansas.

84 Page 81 of Multiple Final Fracture Initiation Sites Multiple Final Fracture Initiation Sites Photograph No. 87 The photographs display overall and close-up views of one of the mating fracture faces from where the final rupture had occurred, resulting in the leakage of crude oil on March 29, The fractographs show the presence of hook cracks adjacent to the fusion line near the O.D. surface along the ERW seam, between a distance of 19' 10" and 21' 6-1/4" from the north girth weld, and radial lines emanating from the ends of the hook cracks as well as chevron marks revealing the crack propagation direction, which is denoted by the arrows.

85 Page 82 of 185 Manufacturing Imperfections Photograph No. 88 The photograph displays the presence of manufacturing imperfections that were found between a distance of 1/4" and 2' 2" from the north girth weld in the path of the final fracture.

86 Page 83 of 185 Photograph No. 89 Photograph No. 90 The photographs display evidence of manufacturing imperfection, i.e. the upturned bands near the O.D. in the fracture path of the final fracture.

87 Page 84 of 185 Photograph No. 91 Photograph No. 92 The photographs display the continuation of the manufacturing imperfections in the path of the final fracture.

88 Page 85 of 185 Photograph No. 93 Photograph No. 94 The photographs display evidence of chevron marks pointing downstream toward the fracture origins. The arrows point to some of the fine chevrons.

89 Page 86 of 185 Photograph No. 95 Photograph No. 96 The photographs display the O.D. and I.D. surfaces of a section of the pipe that was removed between a distance of 18' 10" and 22' as measured from the north girth weld and which contained hook cracks along the ERW seam, from where the final failure initiated on March 29, 2013.

90 Page 87 of 185 Fine chevrons Crack propagation toward north girth weld Photograph No. 97 Fine chevrons Crack propagation Photograph No. 98 The photographs display close-up views of the fracture face between a distance of 18' 10" and 19' 4" from the north girth weld of the pipe section, showing faint evidence of chevrons pointing toward the right (south end) near the fracture origins.

91 Page 88 of 185 Crack propagation Photograph No. 99 Crack propagation Photograph No. 100 The photographs display close-up views of the fracture face between a distance of 19' 4" and 19' 10" from the north girth weld of the pipe section, showing chevrons pointing toward the right (south end) near the fracture origins. The arrow in Photograph No. 100 points to the beginning of the hook cracks.

92 Page 89 of 185 Fracture Zone No. 1 Crack propagation Photograph No. 101 Fracture Zone No. 2 Fracture Zone No. 3 Fracture Zone No. 4 Crack propagation Fracture transition through Weld Flash Photograph No. 102 The photographs display close-up views of the fracture face between a distance of 19' 10" and 20' 4" from the north girth weld, showing radial lines, marked by the blue arrows, which originated from hook cracks through the grain flow or banding formed during manufacturing the ERW seam.

93 Page 90 of 185 Fracture Zone No. 5 Fracture Zone No. 6 Fracture Zone No. 7 Crack propagation Photograph No. 103 The photograph displays a close-up view of the fracture face between a distance of 20' 4" and 20' 8" from the north girth weld, showing vertical radial lines emanating from the hook cracks, which are marked by the blue arrows, indicating the primary fracture initiation sites which resulted in the 22' 3" long fracture along the ERW seam of the 49' 9-1/2" long pipe.

94 Page 91 of 185 Fracture Zone No. 8 Fracture Zone No. 9 Fracture Zone No. 10 Fracture transition through Weld Flash Crack propagation Photograph No. 104 Fracture Zone No. 11 Hook cracks Hook cracks Chevrons Crack propagation Photograph No. 105 The photographs display close-up views of the fracture face between a distance of 20' 8" and 21' 1" from the north girth weld, showing radial lines emanating from the hook cracks, marked by the blue arrow, and chevrons pointing to the cracks, revealing some of the final fracture origins.

95 Page 92 of 185 Fracture Zone No. 12 Crack propagation Photograph No. 106 Fracture Zone No. 13 Fracture Zone No. 14 Crack propagation Photograph No. 107 Crack propagation Photograph No. 108 The photographs display close-up views of the fracture face between a distance of 21' 1" and 21' 10" from the north girth weld showing radial lines emanating from the hook cracks. The blue arrows point to the radial lines, indicative of some of the final fracture initiation sites.

96 Page 93 of 185 Crack propagation Photograph No. 109 Crack propagation Photograph No. 110 The photographs display close-up views of the fracture face between a distance of 21' 10" and 22' from the north girth weld, showing the final fracture which terminated in the base metal of the pipe, diagonally to a distance of approximately 3".

97 Page 94 of 185 Photograph No. 111 Photograph No. 112 The photographs display the O.D. and I.D. surfaces adjacent to one of the mating fracture faces which contained multiple hook cracks. The arrow points to an area where the coating was apparently damaged prior to the incident.

98 Page 95 of 185 Photograph No. 113 Fractured End Photograph No. 114 The photographs of the outside surface of the fractured ERW seam at a distance between 20' 4-1/2" and 20' 6" from the north girth weld show evidence of what appears to be crack or melting caused by copper electrode contacts during the ERW seam fabrication. The arrows point to these imperfections.

99 Page 96 of 185 Photograph No. 115 Photograph No. 116 The photographs display close-up views of the copper electrode contact marks in the heat-affected zone of the ERW seam, at the arrow, on the O.D. surface and the presence of copper.

100 Page 97 of 185 Photograph No. 117 Photograph No. 118 The photographs display the mating fracture faces between a distance of approximately 20' 2-1/2" and 20' 8" from the north girth weld, revealing hook cracks in the heat-affected zone of the ERW seam to a maximum depth of 0.150" as measured from the O.D. surface, and vertical lines emanating from the tips of the hook cracks, indicative of the final fracture origin sites.

101 Page 98 of 185 Hook cracks (discolored area) Some of the final fracture origins Photograph No. 119 Hook cracks (discolored area) Photograph No. 120 The photographs display the mating fracture faces revealing some of the fracture origin site(s) at a distance of approximately 20' 5-5/16" from the north girth weld, which were later examined at higher magnifications using a Scanning Electron Microscope (SEM) to characterize the fracture morphologies.

102 Page 99 of 185 Hook cracks (discolored area) Some of the final fracture origins Photograph No. 121 Hook cracks (discolored area) Photograph No. 122 The photographs display the mating fracture faces revealing some of the fracture origin sites at a distance between 20' 5-3/4" and 20' 7-1/2" from the north girth weld, which were later examined at higher magnifications using an SEM to characterize the fracture morphologies.

103 Page 100 of 185 Hook Crack(s) Final Fracture Photograph No. 123 The SEM fractograph taken of one of the final fracture origin sites at a distance of 20' 5-5/16" from the north girth weld shows an hook crack and the final fracture zone. The fracture locations within the rectangles were examined at high magnifications to further characterize the fracture morphologies. The dotted line denotes the transition zone between the hook cracks and the final fracture.

104 Page 101 of 185 Area-A Photograph No. 124 The SEM fractograph taken of the Area-A of the hook crack near the O.D. surface, as displayed in Photograph No. 123, displays essentially a nondescript featureless fracture surface. Note the absence of any fracture features, likely due to the metal-to-metal contact from the mating fracture faces of the crack and post-crack oxidation. The fracture locations labeled as Location-1A and Location-1B were examined at higher magnifications to further characterize the fracture morphology.

105 Page 102 of 185 Area-A, Location-1A Photograph No. 125 Area-A, Location-1B Photograph No. 126 The SEM fractographs of the two (2) fracture locations labeled as Location-1A and Location-1B in Area-A of the hook crack zone near the O.D. display tightly adhered oxidation product, suggesting that the crack had occurred some time prior to the final fracture.

106 Page 103 of 185 Area-B Photograph No. 127 The SEM fractograph taken of the Area-B of the hook crack zone, as displayed in Photograph No. 123, reveals a nondescript, featureless fracture surface. The fracture location labeled as Location-2A was examined at higher magnification to further characterize the fracture morphology.

107 Page 104 of 185 Area-B, Location-2A Photograph No. 128 The SEM fractograph taken of the Area-B at Location-2A of the hook crack zone, as displayed in Photograph No. 127, reveals tightly adhered oxidation product on the fracture surface.

108 Page 105 of 185 Area-C Photograph No. 129 The SEM fractograph taken of the Area-C of the hook crack zone, as displayed in Photograph No. 123, reveals a nondescript, featureless fracture surface. The fracture locations, labeled as Location-3A, Location-3B, Location-3C, and Location-3D, were examined at higher magnifications to further characterize the fracture morphologies.

109 Page 106 of 185 Area-C, Location-3A Photograph No. 130 Area-C, Location-3B Photograph No. 131 The SEM fractographs taken of the Area-C at Location-3A and Location-3B of the hook crack zone, as displayed in Photograph No. 129, reveal tightly adhered oxidation product.

110 Page 107 of 185 Area-C, Location-3C Photograph No. 132 Area-C, Location-3D Photograph No. 133 The SEM fractographs taken of the Area-C at Location-3C and Location-3D of the hook crack zone, as displayed in Photograph No. 129, reveal tightly adhered oxidation product.

111 Page 108 of 185 Area-D Photograph No. 134 The SEM fractograph taken of the Area-D of the hook crack zone, as displayed in Photograph No. 123, reveals a nondescript, featureless fracture surface. The fracture location within the rectangle was examined at a higher magnifications to characterize the fracture morphology.

112 Page 109 of 185 Area-D within the rectangle Photograph No. 135 The SEM fractograph taken of the Area-D within the rectangle of the hook crack zone, as displayed in Photograph No. 134, reveals tightly adhered oxidation product.

113 Page 110 of 185 Area-E Photograph No. 136 The SEM fractograph taken of the Area-E in the transition zone between the hook crack and the final fracture zones, as displayed in Photograph No. 123, reveals a nondescript, featureless fracture surface. The fracture location labeled as Location-5A was examined at higher magnification to characterize the fracture morphology.

114 Page 111 of 185 Area-E, Location-5A Photograph No. 137 The SEM fractograph taken of the Area-E at Location-5A displays some evidence of oxidation product in the hook crack and evidence of quasi-cleavage separation in the final fracture zone, indicative of pre-existing crack and final brittle fracture, respectively.

115 Page 112 of 185 Hook Crack(s) Final Fracture Area-E, Location-5A, Location within rectangle Photograph No. 138 The SEM fractograph taken of the Area-E at Location-5A, as displayed in Photograph No. 137, confirms the oxidation on the hook cracks and the final fracture in the brittle manner.

116 Page 113 of 185 Area-F Photograph No. 139 The SEM fractograph taken of the Area-F of the final fracture zone, as shown in Photograph No. 123, displays unresolved cleavage separation fracture features and faint evidence of ductile microvoid coalescence.

117 Page 114 of 185 Area-F, Location-6A Photograph No. 140 The SEM fractograph taken of the Area-F at Location-6A of the final fracture zone confirm the presence of predominantly brittle failure with some isolated areas of ductile failure, as indicated by the presence of cleavage separation and patches of microvoid coalescence, respectively.

118 Page 115 of 185 Hook Crack Final Fracture Photograph No. 141 The SEM fractograph taken of the several fracture origin sites at a distance of 20' 6-3/4" from the north girth weld shows an hook crack and the final fracture zone. The fracture areas within the rectangles were examined at higher magnifications to further characterize the fracture morphologies.

119 Page 116 of 185 Area-1 Photograph No. 142 The SEM fractograph taken of the Area-1 of the hook crack fracture zone, as displayed in Photograph No. 141, reveals a highly oxidized fracture surface. The fracture areas, labeled as 1 and 2, were examined at higher magnification to further characterize the fracture morphologies.

120 Page 117 of 185 Area-1, Location-1 Photograph No. 143 Area-1, Location-2 Photograph No. 144 The SEM fractographs taken of the fracture zones labeled as Location-1 and Location-2 in Area-1 of the hook crack reveal a highly oxidized surface and evidence of what appears to be intergranular fracture in a very small fracture zone, respectively. The intergranular fracture may have resulted along the ferrite grain boundaries.

121 Page 118 of 185 Area-2 Photograph No. 145 The SEM fractograph taken of the Area-2 of the hook crack, as displayed in Photograph No. 141, reveals the tightly adhered oxidation product. The area within the rectangle was examined at higher magnification to further characterize the fracture morphology.

122 Page 119 of 185 Area-2, within the rectangle Photograph No. 146 The SEM fractograph taken of the Area-2 within the rectangle in the hook crack, as displayed in Photograph No. 145, reveals a nondescript, featureless fracture surface covered with tightly adhered oxidation product.

123 Page 120 of 185 Area-3 Photograph No. 147 Area-3 Photograph No. 148 The SEM fractographs taken of the Area-3 of the final fracture zone, as displayed in Photograph No. 141, reveal cleavage separation, indicative of brittle failure.

124 Page 121 of 185 Area-4 Photograph No. 149 Area-4 Photograph No. 150 The SEM fractographs taken of the Area-4 of the final shear fracture zone at the I.D. of the pipe reveal evidence of microvoid coalescence, indicative of rapid ductile failure.

125 Page 122 of 185 O.D. Hook Crack(s) Final Fracture I.D. Weld Flash 20' 4-7/8" from the North Girth Weld As-polished, ~25x Photograph No. 151 A composite view of the mating cross-sections removed through the fracture origins area at a distance of 20' 4-7/8" from the north girth weld and prepared for metallographic examination displays evidence of nonmetallic inclusions along the fracture faces and also parallel to the fusion line near the upper half of the pipe wall. Note that the weld flash on the I.D. surface of the pipe was not trimmed off flush with the I.D. surface.

126 Page 123 of ' 4-7/8" from the North Girth Weld As-polished, ~50x Photograph No ' 4-7/8" from the North Girth Weld As-polished, ~50x Photograph No. 153 The micrographs display the upturned inclusions essentially parallel to the fusion line in the ERW upset/haz area, as well as along the fracture faces. Note that vertically aligned inclusions are one of the main contributing factors to the formation of hook cracks.

127 Page 124 of 185 O.D. 20' 4-7/8" from the North Girth Weld As-polished, ~200x Photograph No. 154 Mid-wall 20' 4-7/8" from the North Girth Weld As-polished, ~1000x Photograph No. 155 The micrographs display evidence of folds at the O.D. surface at the fusion line, which was apparently not fully fused, and the presence of post-fracture oxidation at the mid-wall area along the hook crack fracture face.

128 Page 125 of ' 4-7/8" from the North Girth Weld As-polished, ~100x Photograph No. 156 The micrographs display an excessive amount of elongated manganese sulfide inclusions in the diagonal and vertical planes in the upset/haz area of the ERW seam. Note the hook crack along and through the realigned inclusions.

129 Page 126 of ' 4-7/8" from the North Girth Weld As-polished, ~200x Photograph No ' 4-7/8" from the North Girth Weld As-polished, ~200x Photograph No. 158 The micrographs display the manganese sulfide inclusions in the axial direction of the pipe near the I.D. surface of the ERW, which were not affected by the welding process.

130 Page 127 of 185 Hook Crack(s) O.D. Contact Mark Upturned Grains (banding) Primary HAZ Secondary HAZ Final Fracture Fusion Line I.D. Weld Flash 20' 4-7/8" from the North Girth Weld 2% Nital etch, ~20x Photograph No. 159 A composite view of the mating cross-sections removed through the fracture origins area at a distance of 20' 4-7/8" from the north girth weld and prepared for metallographic evaluation shows hook cracks along the brittle upturned bands in the upset/haz area, and the final failure from the tip(s) of the hook crack(s). Again, note that the weld flash was not trimmed off flush with the I.D. surface.

131 Page 128 of 185 O.D. I.D. 20' 4-7/8" from the North Girth Weld 2% Nital etch, ~25x Photograph No. 160 The micrograph displays a hook crack through the upturned bands, which consists of untempered brittle martensite in the upset/haz of the ERW seam.

132 Page 129 of ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 162 The micrographs display the mating fracture faces of hook cracks near the O.D. of the ERW joint. The microstructure consists of grain boundary ferrite and unresolved bainite with some acicular martensite.

133 Page 130 of ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 164 The micrographs display the mating fracture faces of the hook cracks near the mid-wall of the ERW joint. Note the presence of mix-microstructure in the upset/haz of the ERW seam. The upturned bands consist of essentially untempered brittle martensite and the matrix outside of the bands consists of ferrite and unresolved bainite.

134 Page 131 of ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 166 The micrographs display the mating fracture faces of the final crack near the I.D. of the ERW joint. Note the presence of mix-microstructure in the HAZ of the ERW seam consisting of patches of untempered acicular martensite, grain boundary ferrite, and unresolved bainite.

135 Page 132 of ' 4-7/8" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 4-7/8" from the North Girth Weld 2% Nital etch, ~500x Photograph No ' 4-7/8" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 169 The micrographs display the microstructure of the material in the upset/haz between the O.D. and the mid-wall where the upturned bands were formed during the ERW seam manufacturing, consisting of the untempered brittle martensite in the banded area and essentially grain boundary ferrite and unresolved bainite with some patches of untempered martensite in the non-banded area.

136 Page 133 of 185 O.D. Hook Crack(s) Final Fracture Weld Flash I.D. 20' 6-13/16" from the North Girth Weld As-polished, ~25x Photograph No. 170 A composite view of the mating cross-sections removed through the fracture origins area at a distance of 20' 6-13/16" from the north girth weld and prepared for metallographic examination displays evidence of nonmetallic inclusions along the fracture faces, and also parallel to the fusion line near the upper half of the pipe wall. Note that the weld flash on the I.D. surface was not trimmed off flush with the I.D. surface of the pipe.

137 Page 134 of 185 Hook Crack Hook Crack 20' 6-13/16" from the North Girth Weld As-polished, ~50x Photograph No ' 6-13/16" from the North Girth Weld As-polished, ~50x Photograph No. 172 The micrographs display the upturned inclusions essentially parallel to the fusion line in the ERW upset/haz area, as well as along the fracture faces.

138 Page 135 of ' 6-13/16" from the North Girth Weld As-polished, ~200x Photograph No ' 6-13/16" from the North Girth Weld As-polished, ~1000x Photograph No. 174 The micrographs display the presence of several manganese sulfide inclusions aligned parallel to the fusion line and evidence of some post-hook crack oxidation along the fracture face near the mid-wall.

139 Page 136 of ' 6-13/16" from the North Girth Weld As-polished, ~100x Photograph No. 175 The micrographs display an excessive amount of elongated manganese sulfide inclusions aligned in the diagonal and vertical planes in the upset/haz area of the ERW seam. Note the hook crack(s) along and through the realigned inclusions.

140 Page 137 of ' 6-13/16" from the North Girth Weld As-polished, ~200x Photograph No. 176 The micrograph displays the manganese sulfide inclusions in the axial direction of the pipe near the I.D. surface of the ERW, which were not affected by the welding process.

141 Page 138 of 185 Contact Mark Secondary HAZ Primary HAZ Upturned Grains (banding) Fusion Line 20' 6-13/16" from the North Girth Weld 2% Nital etch, ~20x Photograph No. 177 The composite view of the mating cross-sections removed through the fracture origins area at a distance of 20' 6-13/16" from the north girth weld and prepared for metallographic evaluation shows hook crack(s) following the upturned grains and inclusions in the upset/haz area.

142 Page 139 of 185 O.D. Upturned Bands I.D. 20' 6-13/16" from the North Girth Weld 2% Nital etch, ~25x Photograph No. 178 The micrograph displays the hook crack(s) following the upturned bands, which consists of untempered brittle martensite.

143 Page 140 of 185 O.D. O.D. 20' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 180 The micrographs display the mating faces of the hook crack(s) at the O.D. in the ERW seam. The microstructure consists of grain boundary ferrite and unresolved bainite with some acicular martensite.

144 Page 141 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 182 The micrographs display hook crack(s) following the upturned bands of acicular martensite and manganese sulfide inclusions.

145 Page 142 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No. 184 The micrographs display the mating fracture faces of the final fracture near the I.D. of the ERW joint. The microstructure consists of grain boundary ferrite, unresolved bainite, and bands of acicular untempered martensite.

146 Page 143 of 185 Decarburized Surface 20' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 186 The micrographs display the evidence of surface decarburization along the O.D. surface near the ERW seam and the presence of copper from the electrode contact during the initial seam welding of the pipe.

147 Page 144 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 188 The micrographs display the evidence of surface decarburization along the I.D. surface near the ERW seam.

148 Page 145 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 190 The micrographs display one of the contact marks which resulted from the electrical contact between the electrode supplying the welding current and the pipe surface. Note cracks through resolidified metal near the ERW seam within the primary HAZ.

149 Page 146 of 185 Secondary HAZ Upset/HAZ Fusion Line Base Metal Fusion Line to Base Metal 2% Nital etch, ~25x Photograph No. 191 The micrograph displays the microstructural phases between the fusion line and the base metal of the ERW seam.

150 Page 147 of 185 Fusion Line 2% Nital etch, ~100x Photograph No. 192 Fusion Line 2% Nital etch, ~100x Photograph No. 193 The micrographs display untempered bainitic/martensitic microstructure at the fusion line of the ERW seam.

151 Page 148 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 195 The micrographs of the primary HAZ display mix-microstructure consisting of grain boundary ferrite and untempered acicular martensite.

152 Page 149 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 197 The micrographs of the secondary HAZ display essentially the grain boundary ferrite and unresolved pearlite.

153 Page 150 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 199 The micrographs of the base metal display the grain boundary ferrite and lamellar pearlite.

154 Page 151 of ' 6-13/16" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No ' 6-13/16" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 202 The photographs display banded microstructure in the ERW upset area adjacent to the fusion line, consisting of untempered acicular martensite with entrapped ferrite and ferrite with unresolved bainite in the adjacent non-banded matrix.

155 Page 152 of 185 O.D. I.D. 35' 8-1/2" from the North Girth Weld As-polished, ~25x Photograph No. 203 The micrograph of a cross-section removed from the intact ERW seam at a distance of 35' 8-1/2" from the north girth weld displays an excessive amount of manganese sulfide inclusions, some aligned parallel and diagonal to the fusion line during the seam welding process.

156 Page 153 of ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No ' 8-1/2" from the North Girth Weld As-polished, ~500x Photograph No. 205 The micrographs display evidence of some oxidation to a shallow depth of " in the upset/haz.

157 Page 154 of ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No. 207 The micrographs display an excessive amount of manganese sulfide inclusions aligned parallel and diagonal to the fusion line in the upset/haz near the O.D. of the ERW seam joint.

158 Page 155 of ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No. 209 The micrographs display an excessive amount of manganese sulfide inclusions aligned parallel and diagonal to the fusion line in the upset/haz near the mid-wall of the ERW seam joint.

159 Page 156 of ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No. 211 The micrographs display an excessive amount of manganese sulfide inclusions, many of them aligned parallel and diagonal to the fusion line in the upset/haz near the I.D. of the ERW seam joint.

160 Page 157 of ' 8-1/2" from the North Girth Weld As-polished, ~100x Photograph No ' 8-1/2" from the North Girth Weld As-polished, ~500x Photograph No. 213 The micrographs display unfused, expelled weld flash near the I.D. of the ERW seam joint.

161 Page 158 of ' 8-1/2" from the North Girth Weld 2% Nital etch, ~20x Photograph No. 214 The micrograph of the cross-section removed through the intact ERW seam at a distance of 35' 8-1/2" from the north girth weld and prepared for metallographic examination shows upturned as well as downturned bands in the upset/haz, with some bands aligned parallel to the fusion line.

162 Page 159 of ' 8-1/2" from the North Girth Weld 2% Nital etch, ~25x Photograph No. 215 The micrograph displays a composite view of the ERW seam cross-section following etching in a 2% Nital solution revealing some upturned grains parallel to the fusion line.

163 Page 160 of ' 8-1/2" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 8-1/2" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 217 The micrographs display evidence of some oxidation near the O.D. in the upset/haz of the ERW seam joint. The microstructure near the O.D. consists of essentially ferrite and pearlite.

164 Page 161 of ' 8-1/2" from the North Girth Weld 2% Nital etch, ~100x Photograph No ' 8-1/2" from the North Girth Weld 2% Nital etch, ~500x Photograph No ' 8-1/2" from the North Girth Weld 2% Nital etch, ~500x Photograph No. 220 The micrographs display untempered brittle martensite in the bands in the upset/haz of the ERW seam joint.

165 Page 162 of 185 As-polished, ~25x Photograph No % Nital etch, ~25x Photograph No. 222 The micrographs display the microstructural condition at a representative area of the O.D. surface of the pipe, showing the loss of material due to pitting corrosion and the corrosion products adhered to the surface. The insert photograph shows a higher magnification view of a single corrosion pit. The maximum depth of the corrosion pits at this location measured 0.008".

166 Page 163 of 185 As-polished, ~25x Photograph No % Nital etch, ~25x Photograph No. 224 The micrographs display the microstructural condition at a representative area of the I.D. surface of the pipe, showing one of the shallow indentations observed during the visual examination. Note the uniform layer of partial decarburization on the I.D. surface and the grain flow deformation shown in the insert photograph, both indicating that the shallow depression is due to a mechanical indentation, most likely when the pipe was manufactured, and not corrosion pitting. The impression measured 0.137" wide and 0.005" deep.

167 Page 164 of 185 As-polished, ~25x Photograph No % Nital etch, ~25x Photograph No. 226 The composite micrographs display the microstructural condition at another representative area of the I.D. surface of the pipe, showing one of the shallow indentations observed during the visual examination. Note the uniform layer of partial decarburization on the I.D. surface, indicating that the shallow depression is due to a mechanical indentation, most likely when the pipe was manufactured, and not corrosion pitting. The impression measured 0.189" wide and 0.007" deep.

168 Table 1 Page 165 of 185 TO: DIMENSIONAL MEASUREMENTS REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: Measured using a calibrated and certified micrometer IDENTIFICATION: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: CN ' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 Circumference Location of Measurement I.D. Measurement Distance from North Girth Weld Begins Ends -6" 271" 371" 12:00 6: " " " 1:30 7: " " " 3:00 9: " " " 4:30 10: " " " Calculated Out of Roundness 0.111" 0.036" 0.100" th API 5L, 44 Edition, Table 10, Pipe Except End Out-of-Roundness Tolerance for Nominal D = 20" 0.400" THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: Micah Montgomery Laboratory Technician May 8, 2013 M. J. Madhani, Chief Metallurgist E REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-8, REV. 6

169 Table 2 Page 166 of 185 TO: ExxonMobil Pipeline Company SPECIFIED MATERIAL: DIMENSIONAL MEASUREMENTS REPORT th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: DATE OF RECEIPT: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Measured using a calibrated and certified micrometer CN IDENTIFICATION: 33' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 Distance from North Girth Weld Wall Thickness at Crack (inches) Distance from North Girth Weld Wall Thickness at Crack (inches) Distance from North Girth Weld Wall Thickness at Crack (inches) (feet) (inches) West East (feet) (inches) West (East (feet) (inches) West East * *Unable to measure due to geometry of crack tip. TESTED BY: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. Micah Montgomery Laboratory Technician April 24, 2013 M. J. Madhani, Chief Metallurgist E REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-8, REV. 6

170 Table 3 Page 167 of 185 TO: ExxonMobil Pipeline Company SPECIFIED MATERIAL: DIMENSIONAL MEASUREMENTS REPORT th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: DATE OF RECEIPT: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Measured using an Optical Stereomicroscope and calibrated Image Analysis Software CN IDENTIFICATION: 33' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 Distance from North Girth Weld Depth of Cracks Below Surface (inches) Distance from Depth of Cracks Below Surface (inches) Distance from Depth of Cracks Below Surface (inches) O.D. I.D. North Girth Weld O.D. I.D. North Girth Weld O.D. I.D. 19' 10" * * 20' 2" * 20' 6" * 19' 10-1/8" * 20' 2-1/8" * 20' 6-1/8" * 19' 10-1/4" * 20' 2-1/4" * 20' 6-1/4" * 19' 10-3/8" * 20' 2-3/8" * 20' 6-3/8" * 19' 10-1/2" * 20' 2-1/2" * 20' 6-1/2" ' 10-5/8" * 20' 2-5/8" * 20' 6-5/8" ' 10-3/4" * 20' 2-3/4" * 20' 6-3/4" ' 10-7/8" * 20' 2-7/8" * 20' 6-7/8" ' 11" * 20' 3" * 20' 7" ' 11-1/8" * 20' 3-1/8" * 20' 7-1/8" ' 11-1/4" * 20' 3-1/4" * 20' 7-1/4" ' 11-3/8" * 20' 3-3/8" * 20' 7-3/8" * 19' 11-1/2" * 20' 3-1/2" * 20' 7-1/2" * 19' 11-5/8" * 20' 3-5/8" * 20' 7-5/8" * 19' 11-3/4" * 20' 3-3/4" * 20' 7-3/4" * 19' 11-7/8" * 20' 3-7/8" * 20' 7-7/8" * 20' * 20' 4" * 20' 8" * 20' 1/8" * 20' 4-1/8" * 20' 8-1/8" * 20' 1/4" * 20' 4-1/4" * 20' 8-1/4" * 20' 3/8" * 20' 4-3/8" * 20' 8-3/8" * 20' 1/2" * 20' 4-1/2" * 20' 8-1/2" * 20' 5/8" * 20' 4-5/8" * 20' 8-5/8" * 20' 3/4" * 20' 4-3/4" * 20' 8-3/4" * 20' 7/8" * 20' 4-7/8" * 20' 8-7/8" * 20' 1" * 20' 5" * 20' 9" * 20' 1-1/8" * 20' 5-1/8" * 20' 9-1/8" * 20' 1-1/4" * 20' 5-1/4" * 20' 9-1/4" * 29' 1-3/8" * 20' 5-3/8" * 20' 9-3/8" * 20' 1-1/2" * 20' 5-1/2" * 20' 9-1/2" * 20' 1-5/8" * 20' 5-5/8" * 20' 9-5/8" * 20' 1-3/4" * 20' 5-3/4" * 20' 9-3/4" * 20' 1-7/8" * 20' 5-7/8" * 20' 9-7/8" * *No hook cracks at this location. TESTED BY: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. Susan Dalrymple-Ely Materials Analyst April 26, 2013 M. J. Madhani, Chief Metallurgist E REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-8, REV. 6

171 Table 3 (Cont d) Page 168 of 185 TO: ExxonMobil Pipeline Company SPECIFIED MATERIAL: DIMENSIONAL MEASUREMENTS REPORT th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: DATE OF RECEIPT: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Measured using an Optical Stereomicroscope and calibrated Image Analysis Software CN IDENTIFICATION: 33' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 Distance from Depth of Cracks Below Surface (inches) Distance from Depth of Cracks Below Surface (inches) Distance from Depth of Cracks Below Surface (inches) North Girth Weld O.D. I.D. North Girth Weld O.D. I.D. North Girth Weld O.D. I.D. 20' 10" * 21' 3" ' 8" * ' 10-1/8" * 21' 3-1/8" * 21' 8-1/8" * ' 10-1/4" * 21' 3-1/4" * 21' 8-1/4" * ' 10-3/8" ' 3-3/8" * 21' 8-3/8" * ' 10-1/2" ' 3-1/2" ' 8-1/2" * ' 10-5/8" ' 3-5/8" ' 8-5/8" * ' 10-3/4" ' 3-3/4" ' 8-3/4" * ' 10-7/8" ' 3-7/8" ' 8-7/8" * ' 11" * 21' 4" ' 9" * ' 11-1/8" * 21' 4-1/8" * 21' 9-1/8" * ' 11-1/4" * 21' 4-1/4" * 21' 9-1/4" * ' 11-3/8" * 21' 4-3/8" * 21' 9-3/8" * * 20' 11-1/2" * * 21' 4-1/2" * 21' 9-1/2" * * 20' 11-5/8" * * 21' 4-5/8" * 21' 9-5/8" * * 20' 11-3/4" * * 21' 4-3/4" * 21' 9-3/4" * * 20' 11-7/8" * * 21' 4-7/8" * 21' 9-7/8" * * 21' * * 21' 5" * 21' 10" * * 21' 1/8" * * 21' 5-1/8" * 21' 10-1/8" * * 21' 1/4" * * 21' 5-1/4" * 21' 10-1/4" * * 21' 3/8" * 21' 5-3/8" * 21' 10-3/8" * * 21' 1/2" * 21' 5-1/2" * 21' 10-1/2" * * 21' 5/8" * 21' 5-5/8" ' 10-5/8" * * 21' 3/4" * 21' 5-3/4" ' 10-3/4" * * 21' 7/8" * 21' 5-7/8" ' 10-7/8" * * 21' 1" * 21' 6" ' 11" * * 21' 1-1/8" * 21' 6-1/8" ' 11-1/8" * * 21' 1-1/4" * 21' 6-1/4" ' 11-1/4" * * 21' 1-3/8" * 21' 6-3/8" ' 11-3/8" * * 21' 1-1/2" * 21' 6-1/2" * 21' 11-1/2" * * 21' 1-5/8" * 21' 6-5/8" * 21' 11-5/8" * * 21' 1-3/4" * 21' 6-3/4" * 21' 11-3/4" * * 21' 1-7/8" * 21' 6-7/8" * 21' 11-7/8" * * 21' 2" * 21' 7" " * * 21' 2-1/8" ' 7-1/8" ' 2-1/4" ' 7-1/4" ' 2-3/8" ' 7-3/8" ' 2-1/2" ' 7-1/2" ' 2-5/8" ' 7-5/8" ' 2-3/4" ' 7-3/4" ' 2-7/8" ' 7-7/8" * *No hook cracks at this location. TESTED BY: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. Susan Dalrymple-Ely Materials Analyst April 26, 2013 M. J. Madhani, Chief Metallurgist E REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-8, REV. 6

172 Table 4 Page 169 of 185 TO: DIMENSIONAL MEASUREMENTS OF HOOK CRACKS DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: Calibrated Image Analysis Software IDENTIFICATION: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: CN ' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to ' 3-3/4" 20' 4-7/8" 20' 5-1/2" Distance from the North Gird Weld Hook Crack Width Hook Crack Minimum Average Maximum Depth 20' 3-3/4" " " " 0.145" 20' 4-7/8" " " " 0.145" 20' 5-1/2" " " " 0.133" Note: The maximum hook crack depth where measured on the fracture surface was measured to be 0.150", as recorded in Table 3. TESTED BY: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. Clint Myers Staff Metallurgist May 14, 2013 M. J. Madhani, Chief Metallurgist E REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-8, REV. 6

173 Table 5 Page 170 of 185 TO: MICROHARDNESS TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, 44 th Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 SCALE: LOAD FORCE: TEST METHOD: ASTM E INDENTER: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Vickers 500 g Vickers CN IDENTIFICATION: 33' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948; Test Location: 20' 4-7/8" from the North Girth Weld Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale 1 Hardened HRC HRC 2 ERW Upturned HRC HRC Fusion 3 Martensitic HRC 14 Line HRC 4 Grains HRC HRC HRC HRB 6 Hook HRC HRC 7 Crack(s) HRC 18 Secondary HRB HRC 19 HAZ HRC 9 Final HRC HRB 10 Fracture HRC HRB 11 (Primary HAZ) HRC HRB Base HRB Metal HRB Test was performed using calibrated Wilson Tukon Model 230 Tester, S/N Rockwell hardness numbers converted from Knoop or Vickers scales are approximations based on ASTM E and are typically higher than the hardness values obtained using the actual scale. TESTED BY: DATE TESTED: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. J. E. May 10, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-5, REV. 6

174 Table 6 Page 171 of 185 TO: MICROHARDNESS TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, 44 th Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 SCALE: LOAD FORCE: TEST METHOD: ASTM E INDENTER: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Vickers 500 g Vickers CN IDENTIFICATION: 33' 11-1/2" long Fractured Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948; Test Location: 20' 6-13/16" from the North Girth Weld Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale 1 Hardened HRC HRC 2 ERW Upturned HRC HRC Fusion 3 Martensitic HRC 14 Line HRC 4 Grains HRC HRC HRC HRC 6 Hook HRC HRB 7 Crack(s) HRC 18 Secondary HRC HRC 19 HAZ HRB 9 Final HRC HRC 10 Fracture HRC HRB 11 (Primary HAZ) HRC HRB Base HRC Metal HRB Test was performed using calibrated Wilson Tukon Model 230 Tester, S/N Rockwell hardness numbers converted from Knoop or Vickers scales are approximations based on ASTM E and are typically higher than the hardness values obtained using the actual scale. TESTED BY: DATE TESTED: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. J. E. May 10, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-5, REV. 6

175 Table 7 Page 172 of 185 TO: MICROHARDNESS TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, 44 th Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 SCALE: LOAD FORCE: TEST METHOD: ASTM E INDENTER: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: Vickers 500 g Vickers CN IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948; Test Location: 35' 8-1/2" from the North Girth Weld Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale Indentation Number Test Location Hardness, HV500g Conversion to Rockwell Scale 1 Hardened HRC HRC ERW 2 Upturned HRC HRC Fusion 3 Martensitic HRC HRC Line 4 Grains HRC HRC HRC HRB HRC HRC HRC 18 Secondary HRB Primary HRC 19 HAZ HRC HAZ HRC HRB HRC HRB HRC HRB Base HRB Metal HRC Test was performed using calibrated Wilson Tukon Model 230 Tester, S/N Rockwell hardness numbers converted from Knoop or Vickers scales are approximations based on ASTM E and are typically higher than the hardness values obtained using the actual scale. TESTED BY: DATE TESTED: THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. J. E. May 10, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-5, REV. 6

176 Table 8 Page 173 of 185 TO: TENSILE TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: API STD. 5-L, 10 Edition, August 1945, Sections 24-27, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section and Fig. 5b Tested per: ASTM A370-12a ACCEPTANCE CRITERION: th April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: PT ERW API STD. 5-L, 10 Edition, August 1945, Table 3, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, th 44 Edition, October 1, 2007, PSL 1, Table 6, Welded Pipe, Grade X42 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 SAMPLE NUMBER SPECIMEN IDENTIFICATION TEST SPECIMEN DIMENSIONS DIAMETER/ ULTIMATE STRESS YIELD STRESS (0.5% OFFSET) % ELONG. WIDTH, in THICKNESS, in 2 AREA, in LOAD, lbf STRESS, psi LOAD, lbf STRESS, psi IN 2" FRACTURE LOCATION 1 Transverse - ERW , ,000 34,016 77,000 4 H.A.Z. 2 Seam, Weld ,394 93,500 34,938 79,000 5 H.A.Z. 3 Flash Included , ,000 37,194 84, Base Metal 1 Transverse - ERW ,353 85,500 31,104 73,000 3 H.A.Z. 2 Seam, Weld ,341 85,500 31,858 75,000 3 H.A.Z. 3 Flash Removed ,172 92,500 32,440 77,000 5 H.A.Z. REQUIREMENTS th API 5-L, 10 Edition, Table 3, Electric Welded, Open Hearth Steel, Grade B 60,000 minimum REMARKS: th API 5L, 44 Edition, PSL 1, Table 6, Welded Pipe, Grade X42 60,200 minimum Test specimens meet the tensile requirements for API 5L ERW pipe at the time the pipe was manufactured, as well as the current version of API 5L for ERW Pipe, in accordance with the above referenced acceptance criterion. Transverse tensile test specimens were flattened as per API 5L test methods prior to machining and testing. Test was performed using Instron Satec Systems tensile machine S/N THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-3, REV. 7

177 Table 9 Page 174 of 185 TO: TENSILE TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: API STD. 5-L, 10 Edition, August 1945, Sections 24-27, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section and Fig. 5b Tested per: ASTM A370-12a ACCEPTANCE CRITERION: th April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: PT T API STD. 5-L, 10 Edition, August 1945, Table 3, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, th 44 Edition, October 1, 2007, PSL 1, Table 6, Welded Pipe, Grade X42 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 SAMPLE NUMBER SPECIMEN IDENTIFICATION TEST SPECIMEN DIMENSIONS DIAMETER/ ULTIMATE STRESS YIELD STRESS (0.5% OFFSET) % ELONG. WIDTH, in THICKNESS, in 2 AREA, in LOAD, lbf STRESS, psi LOAD, lbf STRESS, psi IN 2" 1 Transverse ,440 87,000 26,343 59, from ,628 86,500 26,288 59, ERW Seam ,329 89,000 27,386 62, FRACTURE LOCATION 1 Transverse ,051 87,000 28,967 63, from ,620 85,500 27,856 60, ERW Seam ,254 87,500 29,443 64, REQUIREMENTS th API 5-L, 10 Edition, Table 3, Electric Welded, Open Hearth Steel, Grade B 60,000 minimum 35,000 min. * th API 5L, 44 Edition, PSL 1, Table 6, Welded Pipe, Grade X42 60,200 minimum 42,100 min. 27 min. REMARKS: Test specimens meet the tensile requirements for API 5L ERW pipe at the time the pipe was manufactured, as well as the current version of API 5L for ERW Pipe, in accordance with the above referenced acceptance criterion. Transverse tensile test specimens were flattened as per API 5L test methods prior to machining and testing. th *The required minimum elongation specified in Table 3 of API STD. 5-L, 10 Edition is illegible on the available paper copy. Test was performed using Instron Satec Systems tensile machine S/N THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-3, REV. 7

178 Table 10 Page 175 of 185 TO: TENSILE TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: API STD. 5-L, 10 Edition, August 1945, Sections 24-27, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section and Fig. 5b Tested per: ASTM A370-12a ACCEPTANCE CRITERION: th April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: PT L API STD. 5-L, 10 Edition, August 1945, Table 3, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, th 44 Edition, October 1, 2007, PSL 1, Table 6, Welded Pipe, Grade X42 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 SAMPLE NUMBER SPECIMEN IDENTIFICATION TEST SPECIMEN DIMENSIONS DIAMETER/ ULTIMATE STRESS YIELD STRESS (0.5% OFFSET) % ELONG. WIDTH, in THICKNESS, in 2 AREA, in LOAD, lbf STRESS, psi LOAD, lbf STRESS, psi IN 2" 1 Longitudinal ,346 89,000 27,764 64, from ,155 90,000 29,107 66, ERW Seam ,043 90,500 30,203 68, FRACTURE LOCATION REQUIREMENTS th API 5-L, 10 Edition, Table 3, Electric Welded, Open Hearth Steel, Grade B 60,000 minimum 35,000 min. * th API 5L, 44 Edition, PSL 1, Table 6, Welded Pipe, Grade X42 60,200 minimum 42,100 min. 27 min. REMARKS: Test specimens meet the tensile requirements for API 5L ERW pipe at the time the pipe was manufactured, as well as the current version of API 5L for ERW Pipe, in accordance with the above referenced acceptance criterion. th *The required minimum elongation specified in Table 3 of API STD. 5-L, 10 Edition is illegible on the available paper copy. Test was performed using Instron Satec Systems tensile machine S/N THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-3, REV. 7

179 Table 11 Page 176 of 185 TO: TENSILE TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company April 16, 2013 SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: API STD. 5-L, 10 Edition, August 1945, Sections 24-27, & th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section and Table 21 Tested per: ASTM A370-12a ACCEPTANCE CRITERION: th P.O. NO.: UCG/ LABORATORY TEST NO.: PT API STD. 5-L, 10 Edition, August 1945, Table 3, Electric Welded, Open Hearth Steel, Grade B, & ANSI/API Spec. 5L, th 44 Edition, October 1, 2007, PSL 1, Table 6, Welded Pipe, Grade X42 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 SPECIMEN IDENTIFICATION Transverse - 90 from ERW Seam Transverse from ERW Seam TEST SPECIMEN DIMENSIONS ULTIMATE STRESS YIELD STRESS (0.5% OFFSET) % DIAMETER/ WIDTH, in THICKNESS, in AREA, in LOAD, lbf STRESS, psi LOAD, lbf STRESS, psi 2 ELONG. IN 2" ,326 86,000 4,169 56, ,492 83,500 4,503 58, REQUIREMENTS % R. IN A. th API 5-L, 10 Edition, Table 3, Electric Welded, Open Hearth Steel, Grade B th API 5L, 44 Edition, PSL 1, Table 6, Welded Pipe, Grade X42 60,000 minimum 60,200 minimum 35,000 min. 42,100 min. th *The required minimum elongation specified in Table 3 of API STD. 5-L, 10 Edition is illegible on the available paper copy. * 27 min. SPECIMEN IDENTIFICATION Transverse - ERW Seam, Weld Flash Removed TEST SPECIMEN DIMENSIONS ULTIMATE STRESS YIELD STRESS (0.5% OFFSET) % DIAMETER/ WIDTH, in THICKNESS, in AREA, in LOAD, lbf STRESS, psi LOAD, lbf STRESS, psi 2 ELONG. IN 2" ,289 99,500 4,765 65,000 21** REQUIREMENTS % R. IN A. th API 5-L, 10 Edition, Table 3, Electric Welded, Open Hearth Steel, Grade B th API 5L, 44 Edition, PSL 1, Table 6, Welded Pipe, Grade X42 **Fractured through the base metal. 60,000 minimum 60,200 minimum Transverse tensile test specimens were not flattened. Test was performed using Instron Satec Systems tensile machine S/N THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 10, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-3, REV. 7

180 Table 12 Page 177 of 185 TO: IMPACT TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 Tested per: ASTM A370-12a ACCEPTANCE CRITERION: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: CI ERW th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 and Table 8, PSL 2 Pipe, Grade X60 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 EFFECTIVE ENERGY: 264 ft-lbf/358 Joules NO. TEST TEMPERATURE SPECIMEN TYPE: V-NOTCH LOCATION Simple Beam, Type A IMPACT VALUES FOR SIZE TESTED, ft-lbf TEST TEMPERATURE: Various LATERAL EXPANSION Plus 95 F ERW Seam Transverse SPECIMEN SIZE TESTED: 10 mm x 5 mm % Shear mils REQUIREMENTS None Specified Plus 80 F ERW Seam Transverse Plus 65 F ERW Seam Transverse Plus 32 F ERW Seam Transverse None Specified None Specified 10 ft-lbf min. average energy 8 ft-lbf min. individual energy Note that the CVN impact requirements are only specified for Type PSL 2 welded pipe, not Type PSL 1 welded pipe. No th impact requirements are listed in the ASI STD 5-L, 10 Edition, August THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-2, REV. 7

181 Table 13 Page 178 of 185 TO: IMPACT TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 Tested per: ASTM A370-12a ACCEPTANCE CRITERION: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: CI HAZ th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 and Table 8, PSL 2 Pipe, Grade X60 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 EFFECTIVE ENERGY: 264 ft-lbf/358 Joules NO. TEST TEMPERATURE SPECIMEN TYPE: V-NOTCH LOCATION Simple Beam, Type A IMPACT VALUES FOR SIZE TESTED, ft-lbf TEST TEMPERATURE: Various LATERAL EXPANSION Plus 95 F ERW Primary HAZ Transverse SPECIMEN SIZE TESTED: 10 mm x 5 mm % Shear mils REQUIREMENTS None Specified Plus 80 F ERW Primary HAZ Transverse Plus 65 F ERW Primary HAZ Transverse Plus 32 F ERW Primary HAZ Transverse None Specified None Specified 10 ft-lbf min. average energy 8 ft-lbf min. individual energy Note that the CVN impact requirements are only specified for Type PSL 2 welded pipe, not Type PSL 1 welded pipe. No th impact requirements are listed in the ASI STD 5-L, 10 Edition, August THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-2, REV. 7

182 Table 14 Page 179 of 185 TO: IMPACT TEST REPORT DATE OF RECEIPT: ExxonMobil Pipeline Company SPECIFIED MATERIAL: th API STD. 5-L, 10 Edition, August 1945, Electric Welded, Open Hearth Steel, Grade B, th & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Welded Pipe, Grade X42 TEST METHOD: th Prepared per: ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 Tested per: ASTM A370-12a ACCEPTANCE CRITERION: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: CI BM th ANSI/API Spec. 5L, 44 Edition, October 1, 2007, Section 9.8 and Table 8, PSL 2 Pipe, Grade X60 IDENTIFICATION: 19' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 EFFECTIVE ENERGY: 264 ft-lbf/358 Joules NO. TEST TEMPERATURE SPECIMEN TYPE: V-NOTCH LOCATION Simple Beam, Type A IMPACT VALUES FOR SIZE TESTED, ft-lbf TEST TEMPERATURE: Various LATERAL EXPANSION Plus 95 F Base Metal Transverse SPECIMEN SIZE TESTED: 10 mm x 5 mm % Shear mils REQUIREMENTS None Specified Plus 80 F Base Metal Transverse Plus 65 F Base Metal Transverse Plus 32 F Base Metal Transverse F Base Metal Transverse Minus 32 F None Specified None Specified 10 ft-lbf min. average energy 8 ft-lbf min. individual energy None Specified Note that the CVN impact requirements are only specified for Type PSL 2 welded pipe, not Type PSL 1 welded pipe. No th impact requirements are listed in the ASI STD 5-L, 10 Edition, August THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Josh Thomas Laboratory Technician May 1, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-2, REV. 7

183 Table 15 Page 180 of 185 TO: ExxonMobil Pipeline Company SPECIFIED MATERIAL: CHEMICAL ANALYSIS REPORT th API STD. 5-L, 10 Edition, August 1945, Table 2, Electric Welded, Open Hearth Steel, th Grade B, & ANSI/API Spec. 5L, 44 Edition, October 1, 2007, PSL 1, Table 4, Welded Pipe, Grade X42 TEST METHOD: ASTM E IDENTIFICATION: DATE OF RECEIPT: April 16, 2013 P.O. NO.: UCG/ LABORATORY TEST NO.: SP ' 10" long Intact Section of a 20" O.D. x 0.312" wall Pipe; Removed from Milepost in the Conway to Corsicana Pegasus Crude Oil Pipeline after it Failed in Service in Mayflower, Arkansas; Installed in 1947 to 1948 ELEMENT WEIGHT % Sample Tested th API 5-L, 10 Ed., Electric Weld Pipe, Open Hearth Steel, Grade B Spec. th API 5L, 44 Ed., PSL 1, Welded Pipe, Grade X42 Specification 1 Carbon max 0.26 max Manganese to max Phosphorus max max Sulfur max max Silicon <0.01 Chromium <0.01 Nickel 0.04 Molybdenum <0.01 Copper 0.02 Aluminum <0.01 Niobium <0.01 Vanadium <0.01 Titanium < max 0.50 max 0.15 max 0.50 max Iron Base Base Base REMARKS: Material analyzed meets the chemical composition requirement for API 5L ERW pipe at the time the pipe was manufactured. However, it does not meet the above referenced current version of API 5L for ERW pipe, in accordance with the above referenced acceptance criterion. 1 Test performed by HurstLab approved supplier and the results are outside the scope of accreditation for tests listed in A2LA Cert. # and not covered by this accreditation. 2 Analytical range not specified for element. 3 Sum of Niobium + Vanadium + Tantalum = 0.15% maximum Test was performed using Thermo Jarrell Ash AtomComp 81, S/N Optical Emission Spectrometer with Angstrom S-1000 readout and control system. THIS IS TO CERTIFY THAT THE ABOVE ARE THE ACTUAL RESULTS OF THE SUBMITTED SAMPLE(S) PREPARED AND TESTED IN ACCORDANCE WITH THE REQUIREMENTS OF THE APPLICABLE SPECIFICATION(S), THE HMRL Q.A. MANUAL, FIFTH EDITION AND ITS IMPLEMENTING PROCEDURES, AS APPLICABLE. TESTED BY: DATE TESTED: Brad Shepard, Chemist May 3, 2013 Joseph Eskew, C.W.I., Laboratory Services Manager THE REPORTED TEST DATA REFLECTS ONLY THE EVALUATED MATERIAL PROPERTIES OF THE ACTUAL TEST SPECIMENS, AND DOES NOT ADDRESS THE MANUFACTURING PROCESSES OR OTHER POSSIBLE REQUIREMENTS SPECIFIED IN THE ABOVE REFERENCED ACCEPTANCE CRITERION. OUR LETTERS AND REPORTS ARE FOR THE EXCLUSIVE USE OF THE CLIENT TO WHOM THEY ARE ADDRESSED. REPRODUCTION OF THE TEST REPORTS EXCEPT IN FULL, AND THE USE OF OUR NAME, MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. TEST SPECIMENS AND/OR UNUSED SAMPLE MATERIAL WILL BE RETAINED FOR 30 CALENDAR DAYS FROM DATE OF REPORT, EXCEPT BY PRIOR AGREEMENT. RLC H M RL FO RM R-7, REV. 6

184 Table 16 Page 181 of 185 Elt. Line Intensity (c/s) Error 2-sig Conc, wt% Mg Ka Al Ka Si Ka S Ka Cl Ka K Ka Ca Ka Ti Ka Mn Ka Fe Ka kv 15.0 Takeoff Angle 15.0 Elapsed Livetime Total It should be noted that EDS analysis is a semi-quantitative test method and was used due to the extremely small sample size. The data obtained should not be used at face value, but only as comparative relative values only. The EDS analysis was performed by an HMRL approved supplier.

185 Table 17 Page 182 of 185 Elt. Line Intensity (c/s) Error 2-sig Conc, wt.% Mg Ka Al Ka Si Ka S Ka Cl Ka K Ka Ca Ka Ti Ka Mn Ka Fe Ka kv 15.0 Takeoff Angle 15.0 Elapsed Livetime Total It should be noted that EDS analysis is a semi-quantitative test method and was used due to the extremely small sample size. The data obtained should not be used at face value, but only as comparative relative values only. The EDS analysis was performed by an HMRL approved supplier.

186 Table 18 Page 183 of 185 Elt. Line Intensity (c/s) Error 2-sig Conc, wt% Mg Ka Al Ka Si Ka S Ka Cl Ka K Ka Ca Ka Mn Ka Fe Ka kv 15.0 Takeoff Angle 15.0 Elapsed Livetime Total It should be noted that EDS analysis is a semi-quantitative test method and was used due to the extremely small sample size. The data obtained should not be used at face value, but only as comparative relative values only. The EDS analysis was performed by an HMRL approved supplier.

187 Table 19 Page 184 of 185 Elt. Line Intensity (c/s) Error 2-sig Conc, wt.% Mg Ka Al Ka Si Ka S Ka Ag La Total kv 15.0 Takeoff Angle 15.0 Elapsed Livetime It should be noted that EDS analysis is a semi-quantitative test method and was used due to the extremely small sample size. The data obtained should not be used at face value, but only as comparative relative values only. The EDS analysis was performed by an HMRL approved supplier.

188 Table 20 Page 185 of 185 Elt. Line Intensity (c/s) Error 2-sig Conc, wt.% Mg Ka Al Ka Si Ka S Ka K Ka Ti Ka Mn Ka Fe Ka Total kv 15.0 Takeoff Angle 15.0 Elapsed Livetime It should be noted that EDS analysis is a semi-quantitative test method and was used due to the extremely small sample size. The data obtained should not be used at face value, but only as comparative relative values only. The EDS analysis was performed by an HMRL approved supplier.

189 Appendix I

190 Appendix I - Page 1 of 6

191 Appendix I - Page 2 of 6

192 Appendix I - Page 3 of 6

193 Appendix I - Page 4 of 6

194 Appendix I - Page 5 of 6

195 Appendix I - Page 6 of 6

196 Appendix II

197 The photograph displays the pipe sections in the as-received condition with the protective wrapping on the outside surface of the pipe sections that was applied to prevent any damage during transportation. Appendix II - Page 1 of 9

198 The photographs display two (2) perspective views of the pipe section in the as-received condition. Appendix II - Page 2 of 9

199 The photograph displays the pipe section, a drum containing the coating material that was removed in the field prior to sectioning of the cracked pipe and a bag containing possible calcareous deposit. Appendix II - Page 3 of 9

200 The photographs display two (2) pipe sections during the unloading process. There was no evidence of any transportation related damage to the pipe sections. Appendix II - Page 4 of 9

201 Appendix II - Page 5 of 9

202 Appendix II - Page 6 of 9

203 Appendix II - Page 7 of 9

204 Appendix II - Page 8 of 9

205 Appendix II - Page 9 of 9

206 Appendix III

207 Appendix III - Page 1 of 5

208 The photographs display the coating removal that was carried by impacting with steel or composite hammers. Appendix III - Page 2 of 5

209 The photographs display the hand removal process of the coating which remained on the pipe after initial removal with hammer. Appendix III - Page 3 of 5

210 The photograph displays the initial coating removal process. Appendix III - Page 4 of 5

211 The photograph displays the careful hand removal process of the coating adjacent to the crack. Appendix III - Page 5 of 5

212 Appendix IV

213 Appendix IV - Page 1 of 26

214 Appendix IV - Page 2 of 26

215 Appendix IV - Page 3 of 26

216 Appendix IV - Page 4 of 26

217 Appendix IV - Page 5 of 26

218 Appendix IV - Page 6 of 26

219 Appendix IV - Page 7 of 26

220 Appendix IV - Page 8 of 26

221 Appendix IV - Page 9 of 26

222 Appendix IV - Page 10 of 26

223 Appendix IV - Page 11 of 26

224 Appendix IV - Page 12 of 26

225 Appendix IV - Page 13 of 26

226 Appendix IV - Page 14 of 26

227 Appendix IV - Page 15 of 26

228 Appendix IV - Page 16 of 26

229 Appendix IV - Page 17 of 26

230 Appendix IV - Page 18 of 26

231 Appendix IV - Page 19 of 26

232 Appendix IV - Page 20 of 26

233 Appendix IV - Page 21 of 26

234 Appendix IV - Page 22 of 26

235 Appendix IV - Page 23 of 26

236 Appendix IV - Page 24 of 26

237 Appendix IV - Page 25 of 26

238 Appendix IV - Page 26 of 26

239 Appendix V

240 The photographs display an overall view of the area along the ERW seam on the intact 19' 10" long section of the pipe, and a closer view of the area where the ERW seam test specimens were removed from. Appendix V - Page 1 of 4

241 The photographs display an overall view of the area opposite the ERW seam on the intact 19' 10" long section of the pipe, showing where the longitudinal and transverse base metal test specimens were removed from. The insert photograph shows the location of the base metal CVN impact test specimens. Appendix V - Page 2 of 4

242 The photograph displays the test specimens that were removed from the intact 19' 10" long section of pipe, after machining and prior to testing. The various test specimens were machined and tested in accordance with ASTM A370-12a and the applicable sections of each edition of API 5L. Appendix V - Page 3 of 4

243 The photographs display the O.D. and I.D. surface, respectively, at the locations where the cross-sections were removed through the fractured area of the ERW seam and metallographically prepared for microstructural evaluation. Appendix V - Page 4 of 4